5- 1

Chapter 5
Interval Estimation
5.1 Concepts of Interval Estimation
5.2 Confidence Intervals for the Mean of a Normal
Distribution: Population Variance Known
5.3 Large-Sample Confidence Interval for a Population
Mean
5.4 Confidence Intervals for the Mean of a Normal
Distribution: Population Variance Unknown
5.5 Confidence Intervals for the Population Proportion
5.6 Determining the Sample Size

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 2

5.1 Concepts of Interval Estimation

Let  be unknown parameter of the population, we take
a random sample of size n from a population X
X:  X1, X2, …, Xn
Use the sample to determine two statistic ˆ1 and ˆ2
Such that

P ˆ1    ˆ2  1   
Then the interval (ˆ1 , ˆ2 ) is called a 100(1-)%
confidence interval for . The quantity (1   )
is called the level of confidence.
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 3

Explanations of
 
P ˆ1    ˆ2  1  

 The confidence interval contains the parameter 

with probability of (1-).
—  is not a random variable
— C.I. is a random interval which varies from
sample to sample.
 A 95% level of confidence means that approximately
95% of these C.I. would contain , and 5% would not.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 4

Steps to find the confidence interval for 

 To construct a sample function for unknown

parameter  with a certain distribution
 To find a critical value for the given the level of
confidence 1- 
 To determine the confidence interval for  using
inequality


P ˆ1    ˆ2  1   

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 5

5.2 Intervals for the Mean of a Normal

Distribution: Population Variance Known
 Given a Normal Population: XN( ,  2)
 And  2 is known.

 Take a random sample of size n:

 X ( X1, X2, …, Xn)

 Construct a “ Sample Statistic ”:

X 
z ~ N (0,1)
/ n

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 6

P Z  z / 2   1  
 X  
P  z / 2   z / 2   1  
 / n 
   
P  X  z / 2    X  z / 2   1
 n n
Confidence Interval :
 
( X  z / 2 , X  z / 2 )
n n

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 7

The C.I. for the mean presented above is

limited by two requirements:

1. The population distribution must be normal:

XN( ,  2);
2.The population variance 2 must be known;
** When the sample size is large, neither
requirement is important.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 8

5.3 Large-Sample Confidence Interval for

a Population Mean
For Large Sample: n 30
(1) Depends on the Central Limit Theorem;
X ~ N (  ,  / n)
2

(2) the sample standard deviation will provide a

fairly accurate estimate of :
s
C.I. for the Mean will be
s s
( X  z / 2 , X  z / 2 )
n n
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 9

Interval Estimates
If the population standard deviation  is known
or the sample size is greater than 30, we use the z
distribution (standard Normal distribution).
s s Standard Normal Distribution
( X  z , X  z ) 0.4
(1   )
2
n 2
n 0.3

We define z  as the z value

f(z)
0.2

2  
that cuts off a right-tail 0.1
2 2

area of 2 under the 0.0
-5 -4 -3 -2 -1 0 1 2 3 4 5
standard normal curve.  z / 2 Z z / 2

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 10

EXAMPLE 3
The Dean of the Business School wants to estimate
the mean number of hours worked per week by
students. A sample of 49 students showed a mean of
24 hours with a standard deviation of 4 hours. What
is the population mean?

 The value of the population mean is not known.

Our best estimate of this value is the sample mean
of 24 hours. This value is the point estimate.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 11

Example 3 continued
Find the 95 percent confidence interval for the
population mean.
s s
( X  1.96 , X  1.96 )
n n
4 4
 ( 24.00  1.96 , 24  1.96 )
49 49
 ( 22.88, 25.12)

The confidence limits range from 22.88 to 25.12.

About 95 percent of the similarly constructed
intervals include the population parameter.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 12

 The American Management Association wishes to have

information on the mean income of middle managers in
the retail industry. A random sample of 256 managers
reveals a sample mean of $45,420. The standard
deviation of this sample is $2,050. The association
would like to answer the following questions:
 What is the population mean?

 What is a reasonable range of values for the population

mean?
 What do these results mean?

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 13
5.4 Confidence Intervals for the Mean of a Normal
Distribution: Population Variance Unknown
 A new Case:
 The population is normal,
 the 2 is unknown,
 the sample size is small.
 Question: What statistic should be used?

X 
z ~ N (0,1)
 / n
X 
t ~ t ( n  1)
s/ n
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 14

P t  t / 2   1  
 X  
P t / 2   t / 2   1  
 s/ n 
 s s 
P X  t / 2    X  t / 2  1
 n n
Confidence Interval is
s s
( X  t / 2 , X  t / 2 )
n n

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 15

Interval Estimates
Ifthe population standard deviation is unknown
and the sample size is less than 30, we use the t
distribution.
s s
( X  t , X  t )
2 n 2 n

where t  is the value of the t distribution

2

with (n-1) degrees of freedom that cuts off a


tail area of
2 to its right.
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 16

Example. The number of cars that pass through a

intersection on weekday mornings
 There are n  8, x  1500, s  300.
 Construct a 99% confidence interval for  ;
 Solution: d.f. = 7, (1-) = 0.99  = 0.01
t / 2 (7)  3.499
300 300
(1500  3.499 , 1500  3.499 )
8 8
 The 99% C.I. is given by
(1128.88, 1871.13)
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 17

5.5 Confidence Intervals for the Population

Proportion (Large Samples)
Population Proportion: p
Sample Proportion: pˆ
pˆ denotes the proportion of 'successes' in n
independent trials, each with probability of
success p
If n is large, then we have
pˆ ~ N ( p, pq / n)
where q  1  p.
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 18

C. I. for the Population Proportion (Large Sample)

 1- Confidence Interval for the population proportion is

( pˆ  z / 2 pq / n , pˆ  z / 2 pq / n )
 Because p and q are unknown, we can use approximately
the following confidence interval

( pˆ  z / 2 pˆ qˆ / n , pˆ  z / 2 pˆ qˆ / n )

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 19

EXAMPLE 4
A sample of 500 executives who own their own
home revealed 175 planned to sell their homes and
retire to Arizona. Develop a 98% confidence
interval for the proportion of executives that plan
to sell and move to Arizona.

0.35  0.65 0.35  0.65

(0.35  2.33 , 0.35  2.33 )
500 500
 (0.35  0.050, 0.35  0.050)
 (0.300, 0.400)

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 20

Definition. Width of a C.I. for the population mean is


w  2 z / 2
n

 The width of a C.I. for the mean depends on three

factors:
1. The level of confidence (1-): z / 2
2. The standard deviation of population: 
3. The sample size: n (controllable)

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 21

5.6 Determining the Sample Size

 Problem: Estimate the mean annual income

 Known  = $4000
 Requirement of C.I. is E  X    500
 for a fix level of confidence （ 1-  ） = 0.95
 Question: How large a sample size is required?

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 22

Solution
 95% C. I. is
4000 4000
( x  1.96 , x  1.96 )
n n
 For
E  X    500 , we have
 Then : 4000
E  1.96  500
n
2 2
1.96 (4000)
n 2
 245.86
500
 At least 246 observations are needed.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 23
Sample Size for Estimating the Mean

 Recall that a 100(1-)% C.I. for  is given by

X 
z ~ N (0,1)
/ n
 The half Width of C.I. Is
z / 2 z / 2
E  n
n E
 then we obtain
z2 / 2 2
n
E2

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 24

Determining the Sample Size

There are 3 factors that determine the size of a
sample, none of which has any direct relationship to
the size of the population. They are:
 The degree of confidence selected.

 The maximum allowable error.

 The variation in the population.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 25

Sample Size for Estimating the Population Mean

 To find the sample size for a variable:
2
 z s 
n  2 
 E 
 

 where : E is the allowable error, z is the z- value

corresponding to the selected level of confidence, and s
is the sample deviation of the pilot survey.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 26

EXAMPLE 6
A consumer group would like to estimate the mean
monthly electricity charge for a single family house in
July within $5 using a 99 percent level of confidence.
Based on similar studies the standard deviation is
estimated to be $20.00. How large a sample size is
required?

2
 2.58  20 
n   107
 5 

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 27

Sample Size for the Population Proportion

 The formula for determining the sample size in the case
of a proportion is:
2
 z 
n  p(1  p ) 2 
E 
 
 where p is the estimated proportion, based on past
experience or a pilot survey; z is the z value associated
with the degree of confidence selected; E is the
maximum allowable error the researcher will tolerate.

School of Economics and Management, Beijing University of Aero/Astronautics

 5- 28

EXAMPLE 7
 The American Kennel Club wanted to estimate
the proportion of children that have a dog as a pet.
If the club wanted the estimate to be within 3%
of the population proportion, how many children
would they need to contact? Assume a 95% level
of confidence and that the club estimated that
30% of the children have a dog as a pet.
2
 1.96 
n  (0.30 )(0.70 )   897
 0.03 
School of Economics and Management, Beijing University of Aero/Astronautics
 5- 29

Summary: Construct a 100(1-) % Confidence

Interval
1. C.I for the Population Mean
(1) Normal population ,  2 is known;
(2) For large sample: n  30;
(3) Normal population, small sample,  2 is
unknown;
2. Confidence Intervals for a Proportion (large
sample)
3. Determining the sample size

School of Economics and Management, Beijing University of Aero/Astronautics