Chapter 8:

STATISTICAL INTERVALS FOR A

SAMPLE SINGLE
Chapter outline
8.1 Introduction.
8.2 Confidence Interval on the Mean of
a Normal Distribution.
8.4 A Large-Sample Confidence Interval For a Population Proportion.
Learning Objectives
After careful study of this chapter, you should be
able to do the following:
1. Construct confidence intervals on the mean of a
normal distribution, using either the normal
distribution or the t distribution method
2. Construct confidence intervals on the variance
and standard deviation of a normal distribution
3. Construct confidence intervals on a population
proportion
Learning Objectives
4. Use a general method for constructing an
approximate confidence interval on a parameter
5. Construct prediction intervals for a future
observation
6. Construct a tolerance interval for a normal
population.
7. Explain the three types of interval estimates:
confidence intervals, prediction intervals, and
tolerance intervals
 8.1. Introduction
 In the previous chapter we illustrated how a
parameter can be estimated from sample data.
However, it is important to understand how good is
the estimate obtained.
 Bounds that represent an interval of plausible values
for a parameter are an example of an interval
estimate.
 Three types of intervals will be presented:
• Confidence intervals
• Prediction intervals
•
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Suppose that X , Xis ,...,
a random
X sample from a normal distribution with
unknown mean 1 2 n
μ and known variance . From the results of Chapter 6, we know that the
sample mean is
 2
normally distributed with mean μ and variance
X

We may standardize X by subtracting the mean and dividing by the standard
2

deviation, which results in the variable .The random variable n

Z has
a standard normal distribution.

X 
Z
/ n
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
A confidence interval estimate for  is an interval of the
from l  u , where the end-point l and u are computed
from the sample data. Because different samples will
produce different values of l and u , these end-points are
values of random variables L and U , respectively. Suppose
that we can determine values of L and U such that the
following probability statement is true:
P  L  U   1  
where 0  1.There is a probability of 1   of selecting a
sample for which the CI will contain the true value of  .
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Once we have selected the sample, so that X 1  x1 , X 2  x2 ,
..., X n  xn ,and computed l and u , the resulting confidence
interval for  is l  u.
The end-points or bounds l and u are called the lower-and
upper-confidence limits (bounds), respectively, and 1   is
called the confidence coefficient.
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
 
In our problem situation, because Z  X   /  / n has 
a standard distribution, we may write
 X  
P   z /2   z /2   1  
  / n 
   
Now manipulate the quantities inside the brackets by (1)
multiplying through by  / n, (2) subtracting X from each
term, and (3) multiplying through by  1. This result in
   
P   X  z /2   X  z /2   1  
 n n 
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Definition: If x is the sample mean of a random sample of
size n from a normal population with known variance  2
a 100 1   % confidence interval (CI) on  is given by
 
x  z /2   x  z /2
n n

where z /2 is the upper 100 / 2 percentage point of the

standard normal distribution.
 8.2 Confidence Interval on the Mean of
a Normal Distribution.

X 
Set Z  . Then Z  N 0,1
/ n
P  z /2 Z z /2  1   95% 0.95
 0.05   / 2 0.025
P Z z /2  0.975
z /2 1.96

 z /2 z /2
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
One-Sided Confidence Bounds
Definition
A 100 1   % upper-confidence bound for  is

  x  z 
n
and a 100 1   % lower-confidence bound for  is

x  z  
n
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Confidence Level and Precision of Estimation
The length of a confidence interval is a measure of the
precision of estimation.

Figure: Error in estimating  with x .


E z /2
n
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Choice of sample size
If x is used as an estimate of  , we can be 100 1   %
confident that error x   will not exceed a specified
amount E when the sample size is
2
 z /2  
n  
 E 
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Example : ASTM Standard E23 defines standard test
methods for notched bar impact testing of metallic
materials. The Charpy V-notch (CVN) technique
measures impact energy and is often used to determine
whether or not a material experiences a ductile-to-brittle
transition with decreasing temperature. Ten
measurements of impact energy (J ) on specimens of
A238 steel cut at 60ºC are as follows: 64.1, 64.7, 64.5,
64.6, 64.5, 64.3, 64.6, 64.8, 64.2, and 64.3. Assume that
impact energy is normally distributed with σ = 1J.
a/ Find a 95% CI for μ, the mean impact energy ?
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Example 1:
b/ Construct a lower, one-sided 95% confidence
interval for the mean impact energy ?

c/ Determine how many specimens must be tested

to ensure that the 95% CI on μ for A238 steel cut at
60°C has a length of at most 1.0 J ?
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Solution: a/ The required quantities are
x 64.46, 1, n 10
1   0.95  z /2 z0.025 1.96
The resulting  
x  z /2   x  z /2
95% CI is n n
1 1
64.46  1.96  64.46  1.96
10 10
63.84  65.08
Practical Interpretation: Based on the sample data, a
range of highly plausible values for mean impact
energy for A238 steel at 60°C is
63.84 J  65.08 J
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Solution: b/ The required quantities are
x 64.46, 1, n 10
1   0.95  P Z z  0.95  z 1.64
A lower, one-sided 95% confidence interval for the
mean is 
x  z 
n
1
64.46  1.64  
10
63.94 
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Solution:
c/ Because the bound on error in estimation E is
one-half of the length of the CI, so E 0.5
The required sample size is,
2 2
 z /2   1.96 1 
n     15.37
 E   0.5 

and because n must be an integer, the required

sample size is n = 16.
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Large-Sample Confidence Interval on the Mean
The standard deviation σ is unknown and n 30 .

X 
When n is large, the quantity
s/ n
has an approximate standard normal distribution.
s s
Consequently, x  z /2   x  z /2 is a large-sample
n n

confidence interval for  , with confidence level of

approximately 100 1   %
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Example: An article in the 1993 volume of the
Transactions of the American Fisheries Society
reports the results of a study to investigate the
mercury contamination in largemouth bass. A
sample of fish was selected from 53 Florida lakes,
and mercury concentration in the muscle tissue was
measured (ppm). The mercury concentration values
were
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Example: 1.23 1.33 0.04 0.44 1.2 0.27
0.49 0.19 0.83 0.81 0.71 0.5
0.49 1.16 0.05 0.15 0.19 0.77
1.08 0.98 0.63 0.65 0.41 0.73
0.59 0.34 0.34 0.84 0.5 0.34
0.28 0.34 0.75 0.87 0.56 0.17
0.18 0.19 0.04 0.49 1.1 0.16
0.1 0.21 0.86 0.52 0.65 0.27
0.94 0.4 0.43 0.25 0.27
Find an approximate 95% CI on μ ?
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Solution: The required quantities are
x 0.525, s 0.3486, n 53
1   0.95  z /2 z0.025 1.96
The resulting 95% CI is
s s
x  z0.025   x  z0.025
n n
0.3486 0.3486
0.525  1.96  0.525  1.96
53 53
0.4311  0.6189
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Practical Interpretation: This interval is fairly wide
because there is substantial variability in the
mercury concentration measurements. A larger
sample size would have produced a shorter
interval.
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
t Distribution
Let X 1 , X 2 ,..., X n be a random sample from a normal
distribution with unknown mean  and unknown variance
 2 . The random variable
X 
T
s/ n

has a t distribution with degrees of freedom.

 8.2 Confidence Interval on the Mean of a
Normal Distribution.
The t probability density function is
  k  1 / 2  1
f x     k 1/2
,    x  
 k  k / 2    x 2 / k   1
 
where k is the number of degrees of freedom. The mean and
variance of the t distribution are zero and k / k  2 k  2 ,
respectively.
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
t Distribution

Probability density Percentage points of the

functions of several t distribution.
t distributions.
 8.2 Confidence Interval on the Mean of
a Normal Distribution.
Small-Sample Confidence Interval on the Mean
The standard deviation σ is unknown and n  30 .
If x and s are the mean and standard deviation of a random
sample from a normal distribution with unknown variance
 2 , a 100 1   % confidence interval on  is given by
s s
x  t /2,n 1   x  t /2,n 1 1
n n
where t /2,n 1 is the upper 100 / 2 percentage point of the t
distribution with n  1 degrees of freedom.
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
One-sided confidence bounds on the mean of a
normal distribution are also of interest and are easy
to find. Simply use only the appropriate lower or
upper confidence limit from Equation (1) and
replace t /2,n 1 by t ,n 1
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Example: An article in the journal Materials
Engineering (1989, Vol. II, No. 4, pp. 275–281)
describes the results of tensile adhesion tests on 22 U-
700 alloy specimens. The load at specimen failure is as
follows (in megapascals):

19.8 10.1 14.9 7.5 15.4 15.4

15.4 18.5 7.9 12.7 11.9 11.4
11.4 14.1 17.6 16.7 15.8

Let19.5 8.8 13.6

.Find a 95% 11.9
CI on μ ?11.4
t0.025,21 2.080
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Solution: The required quantities are
x 13.71, s 3.55, n 22
1   0.95  t0.025,21 2.080
The resulting 95% CI is
s s
x  t /2,n 1   x  t /2,n 1
n n
3.55 3.55
13.71  2.08  13.71  2.08
22 22
12.14  15.28
 8.2 Confidence Interval on the Mean of a
Normal Distribution.
Practical Interpretation: The CI is fairly wide
because there is a lot of variability in the tensile
adhesion test measurements. A larger sample size
would have led to a shorter interval.
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Normal Approximation for a Binomial Proportion
If n is large, the distribution of

X  np p  p
Z 
np 1  p  p 1  p 
n
is approximately standard normal.
The quantity is called the standard error of the
point estimator p.1  p  / n
p
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
To construct the confidence interval on p, note that
P  z /2 Z z /2  1  
 p  p 
P   z /2  z /2  1  

 p  1  p  / n 



 
P  p  z /2
 
p 1  p

 p  p  z /2

p 1  p 
1  
n n 
 
 
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Approximate Confidence Interval on a Binomial
Proportion
If p is the proportion of observations in a random
sample of size n that belongs to a class of interest, an
approximate 100(1 – α)% confidence interval on the
proportion p of the population that belongs to this
class is

p  z  
p 1  p
 p  p  z  
p 1  p
 /2  /2
Where n n
is the upper α/ 2 percentage point of the
standardznormal
/2 distribution.
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Summary

p  z 
p 1  p
  p p  z 
p 1  p

 /2  /2
Where: n n


p issample
x / n size,
is the number of individuals in the sample with the
n characteristic.
specified
 x is the upper α/ 2 percentage point of the standard normal
distribution.

z /2
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Example 1:In a random sample of 85 automobile
engine crankshaft bearings, 10 have a surface
finish that is rougher than the specifications allow.
Find a 95% two-sided confidence interval for
proportion of bearings in the population that
exceeds the roughness specification ?
n 85 p  x 10 0.12
x 10 n 85
1   0.95 z /2 z0.025 1.96
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Solution

A point estimate of the proportion of bearings in the

population that exceeds the roughness specification
 10
is p  0.12
85
A 95% two-sided confidence interval for p is

p  z 
p 1  p
  p p  z 
p 1  p

0.025 0.025
n n
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Solution

0.12 0.88  0.12 0.88 

0.12  1.96  p 0.12  1.96
which simplifies to
85 85
0.0509
Practical Interpretation: This CI.
is a wide p 0.189 the sample size
Although
does not appear to be small (n = 85), the value
of is fairly small, which leads to a large standard error for
contributing to the wide CI.
p
p
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Choice of Sample Size

p  z 
p 1  p
  p p  z 
p 1  p

 /2  /2
n n

p  E  p  p  E

The sample size for a specified value E is given by

2
 z /2  
n 
 E 
 p 1  
p

 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Choice of Sample Size
An upper bound on n is given by
2
 z /2 
n   0.25 
 E 
 
because p 1  p 0.25 for any p
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Example 2:
Consider the situation in Example 1.
a/ How large a sample is required if we want to be
95% confident that the error in using p to estimate
p is less than 0.05 ?
p 0.12; z
1   0.95 0.025 1.96
2
E  0.05
n ??
 z /2  
n 
 E 
 p 
1  
p  162.27

n 163
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Example 2:
Consider the situation in Example 1.
b/ If we wanted to be at least 95% confident that
our estimate p of the true proportion p was within
0.05 regardless of the value of p
2
1   0.95  z /2 
n   0.25 384.16
E 0.05  E 
n ?? n 385
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
One-Sided Confidence Bounds
The approximate 100(1 – α)% lower and upper
confidence bounds are

p  z 
p 1  p
  pand p  p  z

p 1  p


n n
respectively
 8.4 A Large-Sample Confidence Interval
For a Population Proportion.
Example 3: The fraction of defective integrated
circuits produced in a photolithography process is
being studied. A random sample of 300 circuits is
tested, revealing 13 defectives. Calculate a 95%
upper confidence bound on the fraction of
defective circuits ?
n 300 x 13

z0.05 1.645; p   0.043
x 13 n 300
1   0.95 p 0.0627