SAMPLING

DISTRIBUTIONS
& CONFIDENCE
INTERVAL
CHAPTER 2
BKU2032

1
CONTENT
2.1 Sampling Distribution
2.2 Estimate, Estimation and Estimator
2.3 Confidence Interval for the Mean μ
2.4 Confidence Interval for the Difference
between Two Means
2.5 Confidence Interval for the Proportion
2.6 Confidence Interval for the Difference
between Two Proportions
2.7 Confidence Interval for Variances and
Standard Deviations
2.8 Confidence Interval for Ratio of Two Variances
and Standard Deviations 2
 2.1 Sampling Distributions

OBJECTIVE

 After completing this chapter, you should be able to

1. Identify the sampling distribution for sample mean,

difference between two sample means, sample
proportion and difference between two sample
proportions.

3
 Introduction
 Focus on sampling from distributions or
populations & study the important
characteristics of the populations.

 Sampling distribution of a statistics depends on

 the size of the population
 the size of the samples
 the method of choosing the samples

4
 Sampling Distribution of X
Suppose we have a set of independence random variable
X 1 , X 2 ,..., X n where E  X    and Var ( X )   2 for i  1,2,..., n.
i i
For this random sample, the sampling distribution of a single
mean is
  2

X ~ N   , 
 n 
Notes:
• If the population is normally distributed, the sampling distribution is
normal regardless of sample size.
• By using the Central Limit Theorem, if the population distribution is not
necessarily normal, and has mean μ and standard deviation σ , then, for
sufficiently large n  n  30 , the sampling distribution of X is 
approximately normal, with mean  X   and standard deviation  X 
n
X  X
• The relationship between X and Z : Z 
X
nX 5
 Sampling Distribution for Difference
between Two Means
  12    22 
For X 1 ~ N  1 ,  and X 2 ~ N  2 , 
 n1   n2 

  12  2 2 
X 1  X 2 ~ N  1  2 ,  
 n1 n2 

  12  2 2 
X 1  X 2 ~ N  1  2 ,  
 n1 n2 

Z
 X  X 2    1  2 
Relationship between X 1  X 2 and Z :
1

 12  22

n1 n2 6
Sampling Distribution of Proportion, p̂
 p - population proportion, or Percent of successes for population
 X - sample proportion (percent of successes in a sample) of
pˆ 
n X successes in a sample of size n

 random variable created by counting the number of successes

picked by drawing n times from the population.
X ~ bin  n, p 

 The shape of the binomial distribution looks fairly Normal as long

as n is large and/or p is not too extreme (not close to 0 or 1).
 p 1  p  
If np 1  p   5, then p ~ N  p,
ˆ 
 n 
 Relationship between p̂ and Z: Z  pˆ  p
p 1  p 
n
7
Sampling Distribution for Difference
between Two Proportions
 p1 1  p1    p2 1  p2  
If pˆ1 ~ N  p1 ,  and pˆ 2 ~ N  p2 , 
 n1   n2 

 p1 1  p1  p2 1  p2  
pˆ1  pˆ 2 ~ N  p1  p2 ,  
 n1 n 2 

 p1 1  p1  p2 1  p2  
pˆ1  pˆ 2 ~ N  p1  p2 ,  
 n1 n2 

with Z
 pˆ1  pˆ 2    p1  p2 
p1 1  p1  p2 1  p2 

n1 n2
8
 2.2 Estimate, Estimation,
Estimator
OBJECTIVE

 After completing this chapter, you should be able to

1. Define and understand the general formula of interval

estimate (confidence interval) for a parameter.

9
Estimator

 Probability function are actually families of models

in the sense that each include one or more
parameter.

 Example: Poisson, Binomial, Normal

 Any function of a random sample whose objective

is to approximate a parameter is called a statistic
or an estimator

 ˆ is the estimator for 

statistic parameter
10
Properties of Good Estimator
 Unbiased
–
 Efficient
–

 Sufficient
–

 Consistent
–

11
 Estimations & Estimate
 Estimation – Is the entire process of using an estimator to
produce an estimate of the parameter

 2 types of estimation
1. Point Estimate
• A single number used to estimate a population parameter
Parameter,  Point estimator, ˆ
 X
 2
S2
p p̂

2. Interval Estimate
• A spread of values used to estimate a population
parameter
• The interval is usually written (a, b) where a and b are
known as confidence limit
12
• a – lower confidence limit b – upper confidence limit
Definitions
 Confidence Interval
– Range of numbers that have a high probability
of containing the unknown parameter as an
interior point.
– By looking at the width of a confidence interval,
we can get a good sense of the estimator
precision.
– Width = b – a

 Confidence Coefficient 1

– The probability of correctly including the
population parameter being estimated in the
interval that is produced
13
Definitions
 Level of Confidence
– The confidence coefficient expressed as a
percent ,
– Example: 1    100%

1    %
14
Definitions

OR
(1- α) 100% confidence interval for θ

= ˆ  distribution for ˆs.d. for ˆ

15
 2.3 Confidence Interval for
Mean
OBJECTIVES

 After completing this chapter, you should be able to

1. Find the confidence interval for the mean.

2. Find the confidence interval for the mean when σ is

known and unknown.

16
 Confidence Interval for the Mean

The ( 1 – α ) 100 % confidence interval for μ

   
 X  z , X  z 
 2 n 2 n
 s s   s s 
 X  z , X  z   X  t , n 1 , X  t , n 1 
 2 n 2 n  2 n 2 n

17
t- Distributions

The number of values that are free

to vary after a sample statistic has
been computed

18
Rounding Rule
 When you are computing a confidence
interval for a population mean by using raw
data, round off to one more decimal place
than the number of decimal places in the
original data.

 When you are computing a confidence

interval for a population mean by using a
sample mean and standard deviation, round
off to the same number of decimal places
as given for the mean.

19
 Example 2.1
Bonding strength of hydrocarbon is normally
distributed with a standard deviation of 2.18039 105 .
A sample of 16 specimens has been randomly
selected which gives the sample mean of 2.49978  10 7
.
Construct a 95% confidence interval on the mean
bonding strength.

Answer: 2.4891107 ,2.5105 107 

20 20
 Example 2.2
A random sample of 35 items was selected and the
specified brightness was evaluated for each item
by measuring the amount of current required. The
sample mean of current required is 280.3
microamps and the sample standard deviation is
10.3 microamps. Construct a 90% confidence
interval on the mean current required.

Answer: (277.44, 283.16) microamps

21 21
 Example 2.3
A random sample of 12 wafers was drawn from a
slider fabrication process which gives the following
photoresist thickness in micrometer:
10 11 9 8 10 10 11 8 9 10 11 12
Assume that the thickness is normally distributed.
Construct a 95% confidence interval for mean of all
wafers thickness produced by this factory.

Answer: (9.1, 10.7)

22 22
 Example 2.4

How large a sample is required in Example 2.1 if

we want to be 95% confident that our estimate of 
is within ±0.1?

23
 2.4 Confidence Interval for the
Difference Between 2 Means

OBJECTIVES

 After completing this chapter, you should be able to

1. Find the confidence interval for the difference between

two means when σ’s are known.

2. Find the confidence interval for the difference between

two means when σ’s are unknown and equal.

3. Find the confidence interval for the difference between

two means when σ’s are unknown and not equal. 24
Confidence Interval for the
Difference Between 2 Means

25
 Example 2.5
In a random samples of 49 Silver Tyres and 36 Dun
Tyre were selected. The sample mean mileage the
tyre lasts for Silver Tyres is 119 000 km and the
standard deviation is 7700 km, meanwhile the
sample mean mileage the tyre lasts for Dun Tyres is
118 000 km and the standard deviation is 6000 km.
Construct a 90% confidence interval of the two
population means.

Answer: ( -1445, 3445) km

26 26
 Example 2.6
A process engineer is comparing two different
etching solution for removing silicon from the
backs of wafers. The etch rates are follow the
normal distribution and have the same population
variances. Below are the observed each rates from
10 wafers for each solution.
Solution 1 9.7 9.3 9.1 9.5 10.0 10.5 10.2 9.9 10.3 10.1

Solution 2 10.1 10.5 10.6 10.3 10.3 9.9 10.1 10.2 10.3 10.1

Find a 90% CI for the difference in mean etch rates.

27 27
 Example 2.7
A biotechnology company produces a therapeutic drug whose
concentration is normally distributed. Two new methods of
producing this drug have been proposed, and the method that
produces more drugs will be chosen by management. The researchers
chose a sample of size 15 from each method and obtained the
following data in grams per liter.

Mean Standard deviation

Method X 16.22 0.05
Method Y 17.41 1.00

By assuming that the both population variances are equal, construct a

99% confidence interval on the difference between mean of the
concentration of the both methods.

28
Answer: (-1.90,-0.48) or (0.48,1.90)
 2.5 Confidence Interval
for the Proportion
OBJECTIVE

 After completing this chapter, you should be able to

1. Find the confidence interval for a proportion

29
 The ( 1 – α ) 100 % confidence interval for proportion p

 pq pq 
  p  z , p  z 
 2 n 2 n 

NOTES:
Number of successes in a given sample size
Sample
proportion X
p and q  1  p
n
Sample size
where npˆ  5 and nqˆ  5

30
 Example 2.8
1) The fraction of defective integrated circuits produced in a
photolithography process is being studied. A random sample
of 200 circuits is tested, revealing 13 defectives.
i) Calculate a 95% confidence interval on the fraction of
defective circuits produced by this particular tool.
ii) How large the sample be if we wish to be at least 95%
confident that the error in estimating p is less than 0.02,
regardless of the true value of p?

2) A study is to be conducted of the percentage of home-owners

who own at least two television sets. How large a sample is
required if we wish to be 99% confident that the error in
estimating this quantity is within 5%.

31
2.6 Confidence Interval for the
Difference between
Two Proportions

OBJECTIVE
 After completing this chapter, you should be able to

1. Find the confidence interval for the difference between

two proportions.

32
The ( 1 – α ) 100 % confidence interval for the difference
proportions p1 – p2

 p1q1 p2 q2 
  p1  p2   z  
 2 n1 n2 

33
 Example 2.9
Two different types of injection-molding
machines are used to form plastic parts. A
part is considered defective if it has excessive
shrinkage or is discolored. Two random
samples, each of size 300, are selected, and
15 defective parts are found in the sample
from machine 1 while 8 defective parts are
found in the sample from machine 2.
Construct a 95% confidence interval on the
difference in the two proportions defective.

34
 2.7 Confidence Interval
for Variances and
Standard Deviation

OBJECTIVE
 After completing this chapter, you should be able to

1. Find the confidence interval for a variance and a

standard deviation.

35
The ( 1 – α ) 100 % confidence interval for the variance,  2

  n  1 s 2  n  1 s 2 
 , 2 
  ,n 1 1 ,n 1 
2
 2 2 

The ( 1 – α ) 100 % confidence interval for the standard deviation, 


  n  1 s 2 ,  n  1 s 2 
 2 12 
 2
, n 1
2
, n 1


Where
 
n  1 s 2

~  n21 (Chi-square distribution)

2 36
 Example 2.10

A random sample of 10 rulers produced by a

machine gives a set of data below (in cm).

100.13, 100.07, 100.02, 99.99, 99.88, 100.14,

100.03, 100.10, 99.92, 100.21

Find the 95% confidence interval for the

variance and standard deviation of all the
rulers produced by the machine.

37
 Example 2.11
A factory has a machine that’s designed to filled
boxes with an average of 24 ounces of cereal, and
the population standard deviation for this filling
process is expected to be 0.1 ounce.
Thus, if the machine is working properly, the
population variance should be 0.01 squared ounce.
To estimate the value of population variance, an
employee selected a random sample of 15 boxes
from a supply filled by the machine and found that
the sample variance was 0.008 squared ounce.
What is the 98% confidence interval for the
population variance and standard deviation?

38
 2.8 Confidence Interval
for the Ratio of Two Variances
and
Standard Deviations
OBJECTIVE
 After completing this chapter, you should be able to

1. Find the confidence interval for the variance and a

standard deviation proportion.

39
 The ( 1 – α ) 100 % confidence interval
 12
for the ratio of variance, 2
2

 2 2 
 s1 1 s1
 2 , 2 F ,n 1,n 1 
 s2 F ,n 1,n 1 s2 2 2 1 
 2 1 2 

Where
s12  12
~ Fn1 1,n2 1 (F distribution)
s2  2
2 2

1
F1 , n1 1, n2 1

2 F , n2 1, n1 1
2
40
Example 2.12
The machine in Example 2.10 is serviced. A
random sample of 12 rulers produced by the
machine after the serviced made give a set
of data below.
100.03, 100.01, 100.02, 100.04,
99.90, 99.96, 100.04, 100.06,
100.08, 99.98, 100.11, 100.05
Find the 95% confidence interval for ratio of
variance for all rulers produced by the
machine before and after the service.

41
 Conclusion
 An important aspect of inferential statistics is
estimation

 Estimations of parameters of populations are

accomplished by selecting a random sample from
that population and choosing and computing a
statistic that is the best estimator of the parameter

 Statisticians prefer to use the interval estimate

rather than point estimate because they can be
95%,99% or else confidence that their estimate
contains the true parameter and also determine the
minimum sample size necessary.

Thank You
NEXT: CHAPTER 3 HYPOTHESIS TESTING 42