Anda di halaman 1dari 42

Chapter 4: Estimation

CHAPTER 4 : ESTIMATION

Sub-Topic
Introduction.
Point estimation
Interval estimation.
Confidence interval for population mean.
Confidence interval for a difference between two means.
Confidence interval for a population variance.
Confidence interval for ratio of two variances.

Chapter Learning Outcome


Estimate the confidence interval for the single, two population means,
population variance and ratio of two variances.

Learning Objective
By the end of this chapter, students should be able to
Know how to construct confidence interval for population means and
variances.
Able to choose which distribution should be used in order to construct
confidence interval.
Able to choose appropriate sample size.

Key Term (English to Bahasa Melayu)


English

Bahasa Melayu

1.

Point estimation

Penganggar titik

2.

Interval estimation

Penganggar selang

3.

The confidence level

Aras keyakinan

4.

Confidence coefficient

Pekali keyakinan

153

Chapter 4: Estimation

4.1 Introduction

One of the major applications of statistics is estimating population parameters from


sample statistics. If we have an unknown parameter, we may find an estimator for this
parameter and use for the parameter. However, how reliable this estimate is we do not
know. This is where confidence intervals come in. Instead of estimating the
parameter, we say that there is a 95% (or some other percentage) chance that a given
interval contains the parameter. As an example of a parameter estimation problem,
suppose that structural engineer is analyzing the tensile strength is naturally present
between the individual components because of differences in raw material batches,
manufacturing processes and measurement procedures, the engineer is interested in
estimating the mean tensile strength of the components. Knowledge of the statistical
sampling properties of the estimator used would enable the engineer to establish the
precision of the estimate.

4.2 Point Estimation

Definition 1
A point estimate is a single numerical value that used to estimates an unknown
population parameter.
Example 1
For example, a poll may seek to estimate the proportion of adult residents of a city
that support a proposition to build a new sports stadium. Out of a random sample of
200 people, 106 say they support the proposition. Thus in the sample, 0.53 of the
people supported the proposition. This value of 0.53 is called a point estimate of the
population proportion. It is called a point estimate because the estimate consists of a
single value or point.

154

Chapter 4: Estimation

Example 2
Suppose a random variable X is normally distributed with unknown population mean,

. After the sample has been selected, the numerical value of x is the point estimate
of . Thus, if the data are x1 25, x2 30, x3 29, x4 33, the point estimate of is
x

25 30 29 33
29.25
4

4.3 Interval estimation

Definition 2
A confidence interval is a set of (real) numbers between two values that likely to
contain the parameter being estimated.

Definition 3
The confidence level of an interval estimate of a parameter is the probability that the
interval estimates will contain the estimated parameter. Point estimates are usually
supplemented by confidence intervals.

Theory 1
A confidence interval estimate for is an interval of the form l u . There is a
probability of (1 ) of constructing interval that will contain the true value of .
where
P{ l u }= 1
The relationship between and the confidence level is that the stated confidence
level is percentage equivalent to the decimal value of (1 ) .

Definition 4
The end-points or bounds l and u are called the lower-confidence limit and upperconfidence limits respectively.

155

Chapter 4: Estimation

Definition 5
(1 ) is called confidence coefficient.

Example 3
With 95% confidence interval, then 0.05 , since 1 0.05 0.95 .
When 0.01 , then 1 0.01 0.99 , and the 99% confidence interval is being
calculated.

Example 4
If the pollster used a method that contains the parameter 95% of the time it is used, he
or she would arrive at the following 95% confidence interval: 0.46 < < 0.60. The
pollster would then conclude that somewhere between 0.46 and 0.60 of the population
supports the proposal. The media usually reports this type of result by saying that
53% favor the proposition with a margin of error of 7%.

4.4 Confidence Interval for Single Mean


4.4.1 Large Sample : n 30 or known

Theory 2
If the random variable X has a normal distribution with mean and variance 2 ,
then the sample mean X also has a normal distribution with mean , but with
variance 2 / n (refer previous chapter). In other words, X ~ N (

2
n

) . In fact, if we

have a random variable X which has any distribution (not necessarily normal), by the
central limit theorem the distribution of X will be approximately normal with mean

and with variance 2 / n , for large n i.e n 30 .


So standardizing this, we get :

156

Chapter 4: Estimation

2 /n

(since the mean is and the standard deviation is ( / n )


From the Normal Distribution section, we know that P(-1.96 < Z < 1.96) = 0.95.
x

Hence, P 1.96
1.96 0.95
/ n

Rearranging this we get P x 1.96 / n x 1.96 / n 0.95 .


Thus, the 95% confidence interval for is :

x 1.96 / n x 1.96 / n
In general form given as :

x z / n x z / n
/2

(4.1)

/2

x z / 2 / n

or

(4.2)

If the value of is unknown or not given, so the above formula is used by substitute

with s (sample standard deviation). Hence

x z / 2 s / n

(4.3)

The term z / 2 / n is called the maximum error of estimate, E. For specific value,
if 0.05 , 95% of the sample means will fall within this error value on either side
of the population mean. Refer Figure 1.

95%

0.025

z / 2 / n

Figure 1

157

z / 2 / n

0.025

Chapter 4: Estimation

Definition 6
The maximum error of estimate is the maximum likely difference between the point
estimate of a parameter and the actual value of the parameter.

Example 5
Suppose that 100 samples of water from a fresh water lake are taken and the calcium
concentration (milligrams per liter) is measured. The average is 0.66 mg/l and the
standard deviation is 0.049 mg/l. Construct 95% confidence interval for the
population mean.

Answer Example 5
n 100,

x 0.66, s 0.049

1 0.95 0.05
z / 2 z 0.05 / 2 z 0.025 1.96

x z / 2 s / n

0.66 1.96 0.049 / 100

0.66 0.0096
0.6504 0.6696

That is, based on the sample data, a range of the population mean of calcium
concentration from fresh water lake is between 0.6504 and 0.6696.

Theory 3
Sample size determination is closely related to statistical estimation. Quite often, one
asks, how large a sample is necessary to make an accurate estimate ? To determine
the minimum sample size for finding a confidence interval for the mean, the formula
for sample size is derived from the maximum error of estimate formula.

E Z 2

(4.4)

158

Chapter 4: Estimation

and this formula is solved for n as follows.

E n Z 2

Hence,

Z 2
E

Z 2

n
E

(4.5)

From (4.2) x z / 2 / n . Since E z / 2 / n , hence confidence interval for


also can be written as x E .

Example 6
A manufacturer is interested in the output voltage of a power supply used in a PC.
Output voltage is assumed to be normally distributed, with standard deviation 1.25V.
How large a sample must be selected if he wants to be 99% confident of finding
whether the true mean differs from the sample mean by 0.3V ?

Answer Example 6

1.25, 1 0.99 0.01, E 0.3


z / 2 z 0.01/ 2 z 0.005 2.5758

z
n /2

2.5758(1.25)

0.3

115.1866
116

159

Chapter 4: Estimation

4.4.2 Small Sample: n 30 and Unknown

Theory 4
When 2 is unknown, a logical procedure is replace 2 with the sample variance s.
The random variable Z now becomes T

x
s2 / n

, whereby the random variable has a

t distribution with n 1 degree of freedom. Hence the formula for mean population is
given by:
s
x t 2 , v

__

(4.6)

with degree of freedom, v n 1 .

Example 7
The comprehensive strength of concrete is being tested by civil engineer. He tests 10
specimens and the obtained data presented in Table 1.
Table 1
2590

2530

2510

2566

2541

2557

2582

2550

2583

2599

Construct a 98% confidence interval for mean strength of the concrete.

Answer Example 7
x 2560 .8, s 28.5455, n 10

1 0.98 0.02
t / 2,v t 0.02 / 2,101 t 0.01,9 2.821


2560 .8 2.82128.5455 /
x t / 2,v s / n

10

2560 .8 25.4648
2535 .3352 2586 .2648

160

Chapter 4: Estimation

That is, based on the sample data, a range of the population mean strength of the
concrete is between 2535 .3352 and 2586 .2648 .

Example 8
A civil engineer is analyzing the compressive strength of concrete. A random sample
of 12 specimens has a mean compressive strength of 3201.33 psi and a standard
deviation of 900 psi. Construct a 95% confidence interval for the mean compressive
strength.

Answer Example 8
x 3201 .33, s 900, n 12

1 0.95 0.05
t / 2,v t0.05 / 2,121 t0.025,11 2.201

x t / 2,v s / n

3201.33 2.201 90 / 12

3201.33 57.1837

3144 .1463 3258 .5137


That is, based on the 12 sample, we are 95% confident that the mean compressive
strength is between 3144.1463 and 3258.5137 psi.
Exercise 4.4
Objective Questions
1.

A 95% confidence interval estimate can be interpreted to mean that :


(a)

If all possible samples are taken and confidence interval estimates are
developed, 95% of them would include the true population mean
somewhere within their interval.

(b)

You have 95% confidence that you have selected a sample whose
interval does include the population mean.

(c)

b is true

(d)

both a and b are true

161

Chapter 4: Estimation

2.

Which of the following statements is false ?


(a)

There is a different critical value for each level of alpha.

(b)

You can construct a 100% confidence interval estimate of .

(c)

Alpha is a proportion in the tails of the distribution that is outside the


confidence interval.

(d)

In practice, the population mean is the unknown quantity that is to be


estimated.

3.

In the construction of confidence intervals, if all other quantities are


unchanged, an increase in the sample size will lead to ________interval.

4.

(a)

a narrower.

(b)

a wider.

(c)

the same.

(d)

a less significant.

Other things being equal, as the confidence level for a confidence interval
increase, the width of the interval increases.

5.

(a)

True.

(b)

False.

The t distribution is used to construct confidence intervals for the population


mean when
(a)

the population standard deviation is known and the sample size is


greater than 30.

(b)

the population standard deviation is known or the sample size is


greater than 30.

(c)

the population standard deviation is unknown and the sample size is


small.

(d)

the population standard deviation is unknown or the sample size is


small.

162

Chapter 4: Estimation

Subjective Questions
6.

The mean yield of a chemical process is being studied by an engineer. From


previous experience with this process the standard deviation of yield is known
to be 3. He would like to be 99% confident that the estimate should be
accurate within yield with the value of one.
(a)

Determine the error.

(b)

How large a sample is necessary for this study ?

(c)

If it was found that the sample mean is 10, find a 99% confidence
interval for the mean yield.

7.

In the production of airbag inflators for automotive safety systems, a company


is interested to estimate the true mean of the inflator. Measurements on 20
inflators yielded an average value 2.02 cm and a standard deviation of 0.05.
Find a 98% confidence interval of the true mean.

8.

Suppose that in question 7 exercise 4.4 the sample size has been increased
from 20 to 50 inflators. Given that other measurements are still the same. Is
there any change for the confidence interval of the true mean ? If the answer is
yes, find the new confidence interval of the true mean.

9.

The mean breaking strength of yarn used in manufacturing drapery material is


being studied. Past experience has indicated that the standard deviation of
breaking strength is 3.1 psi. A random sample of nine specimens is tested, and
the average breaking strength is found to be 99.4 psi. Find a 95% confidence
interval of the mean breaking strength.

10.

A research engineer for a tire manufacturer is investigated tire life for new
rubber compound. Sample of 18 tires had been tested to end-of-life in a road
test and their average is 63,559 kilometers and standard deviation is 2,275
kilometers. Find a 99% confidence interval of the mean life tire.

163

Chapter 4: Estimation

11.

The college president asks the statistics teacher to estimate the average age of
the students at their college. How large a sample is necessary ? The statistics
teacher would like to be 99% confident that the estimate should be accurate
within one year. From a previous study, the standard deviation of the ages is
known to be 3 years.

12.

An insurance company is trying to estimate the average number of sick days


that full-time food-service workers use per year. A pilot study found the
standard deviation to be 2.5 days. How large a sample must be selected if the
company wants to be 95% confident of getting an interval that contains the
true mean with a maximum error of one day ?

13.

A restaurant owner wishes to find the 99% confidence interval of the true
mean cost of a dry martini. How large should the sample be if she wishes to
be accurate within RM0.10 ? A previous study showed that the standard
deviation of the price was RM0.12.

14.

A health care professional wishes to estimate the birth weights of infants.


How large a sample must she select if she desires to be 90% confident that the
true mean is within six ounces of the sample mean ? The standard deviation of
the birth weights is known to be eight ounces.

15.

Find the sample size needed to estimate the population mean to within one
fifth of a standard deviation with 99% confidence level.

Answer Exercise 4.4


Objective Questions
1.

2.

3.

4.

5.

164

Chapter 4: Estimation

Subjective Questions
6.

(a)

7.

(b)

60

(c)

(9,11)

(1.9916, 2.0484)

8.

(2.0036, 2.0364)

9.

(97.3747, 101.4253)

10.

(62,005.0268, 65,112.9732)

11.

60

12.

25

13.

10

14.

15.

167

4.5 Confidence Interval for a Difference Between Two Means

Theory 5
In the previous chapter we already know that the mean sampling distribution of

12 2 2
.
X 1 X 2 is given by X 1 X 2 ~ N 1 2 ,

n1
n2

In the same way as confidence interval for population mean, we obtained the formula
for confidence interval for the difference between two means. Hence the confidence
interval in general form given by :
2
2
2
2

x1 x2 z / 2 1 2 1 2 x1 x2 z / 2 1 2

n1

n2

n1

n2

(4.7)

2 2
1
2
n1
n2

x1 x2 z / 2

or
(4.8)

We also decide whether to use the Z distribution or t distribution by using the same
principle.
If the sample size is large ( n 30 ) or known, we will use the Z
distribution.
If the sample size is small ( n 30 ) and unknown, we will use the t
distribution.

165

Chapter 4: Estimation

In other words, we only use the t distribution when the sample size is small ( n 30 )
and unknown, for other conditions we will use the Z distribution.
Z distribution case
When the value of is known, we will use (4.8) formula, but when is unknown
we simply substitute with the sample standard deviation, s. Thus from (4.8)
formula, we will obtained:

s2 s 2
1
2
n1 n2

x1 x2 z / 2

(4.9)

t distribution case
The general formula that will be used when we choose t distribution is given by :

s2 s 2
1
2
n1 n2

x1 x2 t / 2,v

(4.10)

Whereby, v n1 n2 2
We already know that the n1 30 , n2 30 and the value of 1 and 2 are unknown.
When we use t distribution, we need to take note on both sample size and the
information on the population standard deviations. The different information on these
will make the (4.10) formula varies. It can be summarize as follows:

Case (i) : n1 n2 , 21 2 2
Since n1 n2 , we can represent both sample size as n . So, n1 n2 n . From (4.10)
formula, we will obtained:

s2 s 2
1
2
n
n

x1 x2 t / 2,v

(4.11)

With v n1 n2 2 .
Or it can be simplify as :

1 2
2
s1 s 2
n

x1 x2 t / 2,v

166

(4.12)

Chapter 4: Estimation

With v 2n 2

Case (ii) : n1 n2 , 21 2 2
Since 21 2 2 , we can substitute both sample variances with pooled estimated
variance, S p2 . So, s 21 s 2 2 S p2 . From (4.10) formula, we will obtained :

S p2 S p2

x1 x2 t / 2,v

n
n

x1 x2 t / 2,v

Thus,

S p2
n

With v n n 2
Or it can be simplify as :
2

x1 x2 t / 2,v S p

(4.13)

With v 2n 2
S p is a pooled estimate of the variance, which given by :

S P2

(n1 1) s12 (n2 1) s22


n1 n2 2

Case (iii) : n1 n2 , 21 2 2
Since 21 2 2 , we can substitute both sample variances with pooled estimated
variance, S p2 . So, s 21 s 2 2 S p2 . From (4.9) formula, we will obtained :

S p2 S p2

x1 x2 t / 2,v

n1 n2

Or it can be simplify as :

167

(4.14)

Chapter 4: Estimation

1
1

n1 n2

x1 x2 t / 2,v S p

(4.15)

With v n1 n2 2
Case (iv) : n1 n2 , 21 2 2
To construct the confidence interval we still use the general formula in (4.10) :

s2 s 2
x1 x2 t / 2,v 1 2
n1 n2

Instead of using v n1 n2 2 , we must use the following formula to calculate the


degree of freedom :
2

s12 s22

n n2
v 12
2
s12
s22


n
1 n2
n1 1 n2 1

Example 9
A consumer organization collected data on two types of automobile batteries, A and
B. Both populations are normally distributed with standard deviations of 1.29 for
batteries A and 0.88 for batteries B. The summary statistics for 40 observations of
each type yielding average mean of 32.25 hours and 29.81 hours for batteries A and
batteries B respectively. Construct 90% confidence interval for difference between
means life hours for batteries A and batteries B.

Answer Example 9

A 1.29, B 0.88,
n A n B 40
x A 32.25,

x B 29.81

Since n A 30 and n B 30 , A2 and B2 are known, so we will use Z- distribution.

168

Chapter 4: Estimation

0.10, / 2 0.05 and z / 2 1.6449 .


By using (4.8) formula :
2
2

x A xB z / 2 A B

nA

nB

1.29 2 0.882

(32.25 29.81) 1.6449

40

40

2.44 0.4061
2.0339 A B 2.8461

We are 90% confident that the difference between means life hours for batteries A
and batteries B is between 2.0339 and 2.8461 hours.

Example 10
The diameter of steel rods manufactured on two different extrusion machines is being
investigated. Two random samples of size n1 15 and n2 15 are selected, and the
sample means are 8.69cm and 8.51cm and sample variances are 0.30 and 0.44,
respectively from Machine 1 and Machine 2. Construct 95% confidence interval for
difference between means diameter of steel rods manufactured by Machine 1 and
Machine 2.

Answer Example 10

s 21 0.30, s 2 2 0.44,
n1 15, n2 15
x1 8.69,

x2 8.51

Since n1 30 and n2 30 , A2 and B2 are not known, so we will use t- distribution.


n1 n2 so this is case (i) for t-distribution, will use (4.11) formula

1 2
2
s1 s 2
n

x1 x2 t / 2,v
With v 2n 2 then, v 2(15) 2 28

169

Chapter 4: Estimation

1 0.95 0.05 , / 2 0.025 and t / 2,28 2.048


Substitute in (4.11) formula :

1
0.30 0.44
15

8.69 8.51 2.048

0.18 0.4548
0.2748 1 2 0.6348

We are 90% confident that the difference between -0.2748 cm and 0.6348 cm.

Example 11
Reconsider the study on means diameter of steel rods in Example 10. If it is given
that the population variances are equal for both machines, construct the new
confidence interval for difference between means diameter of steel rods manufactured
by Machine 1 and Machine 2 at the same level of significant.

Answer Example 11

s 21 0.30, s 2 2 0.44,
n1 15, n2 15
x1 8.69,

x2 8.51

We will still use t- distribution.


Since 21 2 2 this is case (ii) for t-distribution, will use (4.13) formula
2

x1 x1 t / 2,v S p
With v 2n 2 then, v 2(15) 2 28

1 0.95 0.05 , / 2 0.025 and t / 2, 28 2.048


S P2

n1 1s12 n2 1s22
n1 n2 2
(15 1)0.30 (15 1)0.44
15 15 2

170

Chapter 4: Estimation

0.37

S p 0.6083
Substitute in (4.13) formula :

15

8.69 8.51 2.048(0.6083)


0.18 0.4549

0.2749 1 2 0.6349

We are 90% confident that the difference between means diameter of steel rods
manufactured by Machine 1 and Machine 2 is between -0.2749cm and 0.6349cm.

Example 12
Reconsider the study on means diameter of steel rods in Example 10. If it is given
that the population variances are equal for both machines. Meanwhile the sample size
for Machine 1 has been changed from 15 to 10. Construct the new confidence interval
for difference between means diameter of steel rods manufactured by Machine 1 and
Machine 2 at the same level of significant.

Answer Example 12

s 21 0.30, s 2 2 0.44,
n1 10, n2 15
x1 8.69,

x2 8.51

We will still use t- distribution.


Since 21 2 2 and n1 n2 , so this is case (iii) for t-distribution, will use (4.15)
formula

x1 x 2 t / 2,v

1
1

S p

n
n
2
1

With v n1 n2 2 , then v 10 15 2

1 0.95 0.05 , / 2 0.025 and t / 2,23 2.069

171

Chapter 4: Estimation

S P2

n1 1s12 n2 1s22
n1 n2 2
(10 1)0.30 (15 1)0.44
10 15 2

= 0.3852

S p 0.6206
Substitute in (4.15) formula :

15

8.69 8.51 2.069(0.6206 )


0.18 0.4689

0.2889 1 2 0.6489

We are 90% confident that the difference between means diameter of steel rods
manufactured by Machine 1 and Machine 2 is between -0.2889 cm and 0.6489 cm.

Example 13
Reconsider the study on means diameter of steel rods in Example 10. If it is given
that the population variances are not equal for both machines. Meanwhile the sample
size for Machine 1 has been changed from 15 to 10. Construct the new confidence
interval for difference between means diameter of steel rods manufactured by
Machine 1 and Machine 2 at the same level of significant.

Answer Example 13

s 21 0.30, s 2 2 0.44,
n1 10, n2 15
x1 8.69,

x2 8.51

We will still use t- distribution.


Since 21 2 2 and n1 n2 , so this is case (iv) for t-distribution, will use (4.10)
formula

172

Chapter 4: Estimation

s2 s 2
x1 x1 t / 2,v 1 2
n1 n2

Whereby,

2
1

n1 s 22 n 2

s 2 n
n1
2 2
n1 1
n2 1
2
1

0.30 10 0.44 152


0.30 10 2 0.44 152
10 1

15 1

21.7793
22

1 0.95 0.05 , / 2 0.025 and t / 2,22 2.047


Substitute in (4.10) formula :
0.30 0.44

10
15

8.69 8.51 2.047

0.18 0.4986
0.3186 1 2 0.6786

We are 90% confident that the difference between means diameter of steel rods
manufactured by Machine 1 and Machine 2 is between -0.3186cm and 0.6786cm.

Exercise 4.5
1.

For the following problems, define whether the population standard deviations
are known or unknown. Then, choose the appropriate distribution in order to
construct the confidence interval for the difference in mean.
(a)

The burning rates of two different solid-fuel propellants used in


aircrew escape systems are being studied. It is known that both
propellants have approximately the same standard deviation of burning
rate; that is 4cm/second. Two random samples 25 specimens are tested
for both propellants.

(b)

The behavior of a stochastic generator in the presence of external


noise. The number of periods was measured in a sample of 100 trains

173

Chapter 4: Estimation

for each of two different levels of noise voltage, 100mV and 200mV.
For 100mV, the mean number of periods in a train was 7.9 with
variance of 4.7. Meanwhile, for 200mV, the mean number of periods
in a train was 6.5 with variance of 5.1
(c)

A polymer is manufactured in a batch chemical process. Viscosity


measurements are normally made on each batch, and from past
experience the process has indicated that the variability in the process
is fairly stable with standard deviation of 17. Random sample of 15 are
taken to find their viscosity measurements. A process change is made
which involves switching the catalyst used in the process. Following
the process change, 20 batch viscosity measurements are taken.
Assume the process variability is unaffected by the catalyst change.

(d)

The overall distance traveled by a golf ball is tested by hitting the ball
with Iron Byron, a mechanical golfer with a swing that is said emulate
the legendry champion, Byron Nelson. 10 Randomly selected balls of
two different brands are tested and the overall distance measured. The
obtained data presented in Table 2.
Table 2

Brand A

275

286

287

260

277

273

269

282

281

275

Brand B

258

245

275

270

266

274

269

266

273

273

(e)

A researcher wanted to find out the intentions of young students to


enroll in IT courses in the future. He randomly selected fourth, fifth
and sixth graders and recorded how many IT courses they intend to
take. The obtained information presented in Table 3.
Table 3
n

Mean

Standard deviation

Males

190

2.82

1.41

Females

220

2.42

1.30

174

Chapter 4: Estimation

2.

Refer to question 1(e) Exercise 5.5, given that the sample size for both male
and female have been reduced to 21. Construct a 90% confidence interval for
the difference between males and females in mean number of IT courses
planned to be taken in future.

3.

A group of dietitians is investigating a diet-modification program to stimulate


weight loss. 7 volunteers have participate and their weight (in kilograms)
before and after the participation in the program is shown in the Table 4:
Table 4
Individual

Before

60

71

63

59

75

70

68

After

57

69

56

51

72

66

62

Construct a 98% confidence interval for the difference weight before and after
the program.

4.

The usefulness of two different design languages in improving programming


tasks has been studied. 40 expert programmers, who familiar in both
languages, are asked to code a standard function in both languages, and the
time (in seconds) is recorded. For the Design Language 1, the mean time is
255s with standard deviation of 26s and for the Design Language 2, the mean
time is 319s with standard deviation of 17s. Construct a 95% confidence
interval for the difference in mean coding times between Design Language 1
and Design Language 2.

5.

Two types of plastics are suitable for use by an electronics component


manufacturer. The Breaking strength of this plastic has been studied. It is
known that both types have the same standard deviation. The obtained
information from a random sample presented in Table 5.

175

Chapter 4: Estimation

Table 5
Sample Size

Mean

Variance

Type 1

13

173.5

3.7

Type 2

13

160.9

2.9

Construct a 99% confidence interval for the difference in mean breaking


strength of the two types of plastics.

6.

A study to investigate the melting point of two type alloys is conducted by


melting 50 samples alloy Type I and 60 samples alloy Type II. The sample
mean and standard deviation for alloy Type I was 139 C and 27 C , while for
alloy Type II was 155 C and 31 C respectively. Construct a confidence
interval estimate for the difference in mean melting point for the two type
alloys at significant level 0.02.

7.

The deflection temperature under load for two different types of plastic pipe is
normally distributed with standard deviation of 16 C for Type A pipe and

21 C for Type B pipe. Two random samples of 13 pipe specimens from both
pipe types are tested and their means deflection temperatures are 125 C and
97 C respectively for Type A pipe and Type B pipe. Find a 99% confidence

interval for the difference in mean deflection temperatures between the two
types of pipe.

8.

A chemical engineer wish to know the difference mean yield between two
catalysts that have been used in a certain chemical process. Table 6 shows
data of yields from his experiment result.
Table 6
Catalyst X

80

88

76

83

85

85

81

79

75

Catalyst Y

77

85

81

76

81

74

79

80

83

176

Chapter 4: Estimation

Find a 99% confidence interval for the difference between the means yield
between two catalysts.

9.

Researcher conducted a study to determine whether magnets are effective in


treating back pain. Comparison study has been conducted by giving 9
volunteers a treatment a using magnets and the pain was measured before and
after treatment. Pain was measured using the visual analog scale. Before the
treatment the mean pain is 12.4 with a standard deviation of 2.1. Meanwhile
after the treatment the mean pain is 9.5 with a standard deviation of 2.8. Find
a 98% confidence interval for the difference between mean pain before and
after the treatment. Assume the population variances for before and after the
treatment are equal.

10.

A study to see the difference between BMI of men and women was
conducted. A random sample of 11 men yielding average BMI of 28.9 and
standard deviation of 6.4. A random sample of 13 women yielding average
BMI of 26.1 and standard deviation of 4. Find a 99% confidence interval for
the difference between mean BMI for men and women.

Answer Exercise 4.5


1.

(a)

known, Z distribution

(b)

unknown, t distribution

(c)

known, Z distribution

(d)

unknown, t distribution

(e)

unknown, Z distribution

2.

(-0.3048, 1.1048)

3.

(-5.1154, 14.544)

4.

(-73.627, -54.373)

5.

(10.6071, 14.5929)

6.

(-28.8678, -3.1323)

7.

(46.8607, 9.1393)

8.

(-3.6243, 7.1791)

9.

(-0.1135, 5.914)

10.

(-3.6505, 9.2505)

177

Chapter 4: Estimation

4.6 Confidence Interval for Population Variance

4.6.1 Confidence Interval for Single Population Variance

Theory 6
If s 2 is the sample variance from a random sample of n observations from a normal
distribution with unknown population variance 2 , a confidence interval on 2 is
given as :

(n 1) s 2

2 / 2,v

(n 1) s 2

(4.16)

12 / 2,v

Whereby 2 / 2,v and 12 / 2,v are the upper and lower 100 / 2 percentage points of the
chi-square distribution with v n 1 degree of freedom, respectively.
5.6.2 Confidence Interval for Ratio of Two Population Variances
Theory 7
If s12 and s 22 are the sample variances from a random samples of n 1 and n 2
respectively, from two independent normal populations with unknown population
variances 12 and 22 , a confidence interval on the ratio

1 s12
s12
1

f / 2 ,v 2 ,v1
s22 f / 2,v1,v 2 22 s22

2
is given as :
22
1

Whereby f / 2,v 2,v1 and

1
f / 2 ,v1,v 2

(4.17)

are the upper and lower 100 / 2 percentage points of

the F distribution with v1 n1 1 and v2 n2 1 degree of freedom, respectively.

178

Chapter 4: Estimation

Example 14
The life hours of a heating element used in a furnace is known to be approximately
normally distributed. A random sample of 11 heating elements Type A is selected and
found to have sample standard deviation of 9.7 hours. Meanwhile a random sample of
13 heating elements Type B is selected and found to have sample standard deviation
of 11.4 hours. Construct a 95% confidence interval for the ratio population variances

A2
.
B2

Answer Example 14
n A 11, s A2 9.7 2

nB 13, sB2 11.42

v A n A 1 11 1 10
v B nB 1 13 1 12

0.05, / 2 0.025
f / 2,vB,vA f 0.025,12,10 3.62
1
f / 2,vA,vB

1
f 0.025,10,12

1
3.37

1 s A2
s A2
1

f / 2,vB ,vA
sB2 f / 2,vA,vB 22 sB2
2

12 9.7 2
1

3.62
11.4 2 3.37 22 11.4 2
9.7 2

2
0.2148 2 2.6209
2
1

We are 95% confident that the ratio population variances


2.6209.

179

A2
is between 0.2148 and
B2

Chapter 4: Estimation

Example 15
A polymer is manufactured in a batch chemical process. A random sample of 9
viscosity measurements yielding standard deviation of 19. Following the process
change, 8 batch viscosity measurements yielding standard deviation of 16. Construct
a 98% confidence interval for the ratio population variances

12
.
22

Answer Example 15
n1 9, s`21 19 2
n2 8, s`22 16 2
v1 n1 1 9 1 8
v2 n2 1 8 1 7

0.02, / 2 0.01
f / 2,v 2,v1 f 0.01,7,8 6.18
1
f / 2,v1,v 2

1
f 0.01,8,7

1
6.84

1 s12
s12
1

f / 2 , v 2 ,v 1
s22 f / 2,v1,v 2 22 s22
2

19 2 1 12 19 2
6.18

16 2 6.84 22 16 2

0.2062

2
8.7148
22
1

We are 98% confident that the ratio population variances


8.7148

180

A2
is between 0.2062 and
B2

Chapter 4: Estimation

Exercise 4.6
1.

The percentage of titanium in an alloy used in aerospace casting is measured


in 25 randomly selected parts. The sample standard deviation of 4.8
milligrams. Construct 90% confidence interval for the population variance.

2.

A rivet is to be inserted in a hole. A random sample of 30 parts is selected,


and the hole diameter is measured. The sample variance of the holes diameter
measured is 0.01 millimeters. Construct 95% confidence interval for the
population variance.

3.

The brightness of a television picture tube can be evaluated by measuring the


amount of current required to achieve a particular brightness level. A sample
of 20 tubes results in variance of 16.9. Find the 98% confidence interval for
the population variance.

4.

A particular brand of diet margarine was analyzed to determine the level of


polyunsaturated fatty acid (in percentage). A sample of 12 packages resulted
in standard deviation of 1.8. Find the 99% confidence interval for the
population variance.

5.

A post-mix beverage machine is adjusted to release a certain amount of syrup


into a chamber where it mixed with carbonated water. A random sample of 30
beverages was found to have a variance of 0.058 fluid milliliters. Construct a
90% confidence interval for the population variance.

6.

A manufacturer produces crankshafts for an automobile engine. The wear of


crankshaft after 45,000 km is of interest because it is likely to have an impact
warranty claims. A ransom sample of 10 shafts is tested and obtained a
standard deviation of 0.75. Construct a 95% confidence interval for the
population standard deviation.

181

Chapter 4: Estimation

7.

Consider the weight data in question 3 Exercise 5.5. Construct a 98%


confidence interval for the ratio population variances,

8.

2
Before
2
After

Two chemical companies can supply a raw material. The concentration of a


particular element in this material is important. The standard deviation of
concentration in a random sample of 8 batches produced by Company A is 3.9
grams per liter, while for Company B, a random sample of 11 batches yields
4.7 grams per liter. Construct a 90% confidence interval for the ratio
population variances

9.

A2
B2

A fuel-economy study was conducted for two local automobiles, X and Y. One
vehicle of each brand was selected, and the mileage performance was
observed for 9 tanks of fuel in each car. From the study it was found that X
has a variance of 0.41 liter and Y has a variance of 0.34 liter. Construct a 95%
confidence interval for the ratio population variances.

10.

Table 7 shows the data of waiting times (in minutes) of customers at ABC
Bank and XYZ Bank. Construct a 90% confidence interval for the ratio
population variances.
Table 7
ABC Bank

6.1

5.8

7.0

6.4

6.9

7.3

XYZ Bank

6.8

5.4

7.4

7.1

6.6

6.0

182

6.5

Chapter 4: Estimation

11.

Consider the breaking strength of a plastic described in question 5 Exercise


4.5. Construct a 90% confidence interval for the ratio population variances,
2
Type
1
2
Type
2

12.

Consider the deflection temperature under load for two different types of
plastic pipe described in question 6 Exercise 4.5. Compute a 98% confidence
interval for the ratio population variances,

13.

2
TypeA
2
Type
B

Consider the yields data in question 8 Exercise 4.5. Construct a 95%


confidence interval for the ratio population variances,

14.

X2
Y2

Consider the study to determine whether magnets are effective in treating


back pain described in question 9 Exercise 4.5. Compute a 90% confidence
interval for the ratio population standard deviation,

15.

Before
After

Consider the study to see the difference between BMI of men and women in
question 10 Exercise 4.5. Compute a 90% confidence interval for the ratio
population variances,

2
Male
2
Female

Answer Exercise 4.6


1.

(15.1850, 39.9307)

2.

(0.006343, 0.0181)

3.

(8.8724, 42.0673)

4.

(1.3320, 13.6919)

5.

(0.0395, 0.0950)

6.

(0.5158, 1.3693)

7.

(0.0741, 5.3164)

8.

(0.2193, 2.5063)

183

Chapter 4: Estimation

9.

(0.2722, 5.3421)

10.

(0.1037, 2.253)

11.

(0.6051, 4.3788)

12.

(0.1686, 4.9384)

13.

(0.3519, 5.5311)

14.

(0.1635, 1.935)

15.

(0.9309, 7.4496)

EXERCISE CHAPTER 4
1.

A major truck stop has kept extensive records on various transactions with its
customers. If a random sample of 18 of these records shows average sales of
63.84 gallons of diesel fuel with a standard deviation of 2.75 gallons,
construct a 99% confidence intervals for the mean of the population sampled.

2.

The data represent below is a sample number of fires started by candles at


home for the past several years. Find the 95% confidence interval for the
mean of fires started by candles at home each year as in Table 8.
Table 8
5460

3.

5900

6090

6310

7160

8440

9930

A lathe is set to cut steel bars and considered to be in perfect adjustment if the
average length of the bars it cuts is 7 centimeters. A sample of 31 bars is
selected randomly, and the lengths are measured. It is determined that the
average length of the bars in the sample is 7.055 centimeters with a standard
deviation of 0.35 centimeters. Find a 97% confidence interval that the average
length of the bars is in perfect adjustment.

4.

A random sample of 48 days take at a large hospital shows that averages of 38


patients were treated in the emergency room (ER) per day. The standard
deviation of the population is four. Find the 99% confidence interval of the
mean number of ER patients treated each day at the hospital.

184

Chapter 4: Estimation

5.

To find an interval estimate of the number of miles a certain brand tires will
last before tread depth falls below minimal safety threshold, a manufacturer
tests 50 tires under various operating conditions. The recorded data yield the
estimates x 32,460 miles and s 3106 . Find an 80% confidence interval
for the mean number of miles.

6.

A machine produces metal rods used in an automobile suspension system. A


random sample of 10 rods was selected, and diameter was measured. The
resulting data in millimeter is shown in Table 9.
Table 9
8.23

8.30

8.27

8.22

8.29

8.39

8.21

8.38

8.35

8.37

Find the 98% confidence interval of the mean rod diameter.

7.

The thickness of blended cement for waterproofing application has a normal


distribution. 25 samples of blended cement for water proofing application
were selected randomly and have a mean of 1.17 cm and a standard deviation
of 0.32 cm.
(a)

Determine the standard error of estimate if 0.05 .

(b)

Find the 99% confidence interval of the mean thickness of blended


cement for waterproofing application.

(c)

Find the 90% confidence interval of the mean thickness of blended


cement for waterproofing application.

(d)

Is there any difference between the result in question (b) and (c)?
Why ?

8.

Table 10 below shows the results of a mouse-infection experiment in which


14 mice in Group A and 11 mice in Group B received the same challenge dose

185

Chapter 4: Estimation

of bacteria and were then observed daily. If the population variances of both
group are not equal, find the 99% confidence interval for the difference
between the average days of 2 groups of mice that infected by the dose of
bacteria.
Table 10
Mouse
Group

9.

Day of death (post-infection) of individual mouse

2, 2, 3, 3, 3, 3, 4, 4, 5, 6, 7, 7, 8, 9

1, 1, 2, 2, 2, 2, 3, 3, 3, 4, 5

Two similar groups of patients, A and B, consisting of 50 and 100 individuals


respectively, the first was given a new type of sleeping pill and the second
was given a conventional type. The mean number of hours of sleep for
patients in group A was 7.82 with a standard deviation of 0.24h. While the
mean number of hours of sleep for patients in group B was 6.75 with a
standard deviation of 0.30h. Find 95% confidence limits for the difference in
the mean number of hours of sleep induced by the two types of sleeping pills.

10.

A sample of 14 cans of Brand I diet soda gave the mean number of calories of
23 per can with a standard deviation of 3 calories. Another sample of 16 cans
of Brand II diet soda gave the mean number of calories of 25 per can with a
standard deviation of 4 calories. Assume that the calories per can of diet soda
are normally distributed for each of the two brands and that the standard
deviations for the two populations are equal. Find the 99% confidence interval
for 1 2 .

11.

A study was conducted to investigate some effects of physical training.


Sample data are listed below, with all weights given in kilograms.

Pre training : 99 57 62 69 74 77 59 92 70 85 84

186

Chapter 4: Estimation

Post training : 94 57 62 69 66 76 58 88 70 84 83
Construct a 95% confidence interval for the difference between the weights in
pre training and the weights in post training by assuming 12 22 .

12.

Twelve randomly selected mature Jati trees of one variety have a mean height
of 13.8 feet with a standard deviation of 1.2 feet, and 15 randomly selected
mature Jati trees of another variety have a mean height of 12.9 feet with a
standard deviation of 1.5 feet. Assuming that the random samples were
selected from normal population with equal variances, construct a 95%
confidence interval for the difference between the true average height of the
two kinds of Jati trees.

13.

Reconsider question 1 Exercise Chapter 4, find the 99% confidence interval


for the standard deviation population.

14.

A medical researcher wants to determine whether male pulse rates vary more
or less than female pulse rates. The statistics that he found from his research
can be summarized as shows in Table 11:
Table 11
Male

Female

Number of samples : 7

Number of samples : 9

Mean : 69.4

Mean : 76.3

Standard deviation : 11.3

Standard deviation : 12.5

(a)

Construct 95% of confidence interval for the male standard deviation.

(b)

Construct 95% of confidence interval for the ratio variance between


male and female.

15.

Table 12 below shows difference in waiting times (in minutes) of customers at


the Publician Bank, where customers enter a single waiting line and where

187

Chapter 4: Estimation

customers may enter any one of three different lines that have formed at three
teller windows.
Table 12
Single-line
system
Multiple-line
system

(a)

6.5

6.6

6.7

6.8

7.1

7.3

7.4

7.7

7.7

7.7

4.2

5.4

5.8

6.2

6.7

7.7

7.7

8.5

9.3

10.0

Construct a 95% confidence interval for where the customers enter a


single waiting line.

(b)

Construct a 95% confidence interval for where the customers may


enter any one of three different lines.

(c)

Which arrangement seems better : the single-line system or the


multiple-line system? Why?

16.

Ten alloy of Brand A had an average magnesium content of 3.1 mg with a


standard deviation of 0.5 mg. Eight alloy of Brand B had an average
magnesium content of 2.7 mg with a standard deviation of 0.7 mg. Assuming
that the two sets of data are independent random samples from normal
populations with equal variances.
(a)

Construct a 99% confidence interval for population standard deviation


of Brand A.

(b)

17.

Construct a 98% confidence interval for A B .

The thickness of blended cement for waterproofing application has a normal


distribution. 25 samples of blended cement for water proofing application
were selected randomly and have a mean of 1.17 cm and a standard deviation
of 0.32 cm.
(a)

Determine the standard error of estimate if 0.05

(b)

Find the 99% confidence interval of the mean thickness of blended

188

Chapter 4: Estimation

cement for waterproofing application.


(c)

Find the 90% confidence interval of the mean thickness of blended


cement for waterproofing application.

18.

A random sample of the number of farms (in thousands) in various states in


Malaysia is given in Table 13.
23

45

Table 13
10

18

16

29

18

20

39

38

33

(a)

Find the point estimate for the sample variance.

(b)

Construct the 90% confidence interval of variance for the number of


farms.

19.

Two groups of students are given a problem-solving test, and the results are
shown in Table 14. Construct the 95% confidence interval for the ratio of the
standard deviation for the two groups,

1
.
2

Table 14
Group 1 (Finance Major)
Group 2 (Management Major)
Sample size
Variance

20.

13

15.9

11.4

A study was performed to determine whether men and women differ in their
repeatability in assembling components on printed circuit boards. Two
samples of 15 men and 17 women were selected, and each subject assembled
the units. The two sample standard deviations of assembly time were 1.21
minutes and 1.35 minutes.
(a)

Construct 95% confidence interval of variance assembly time for men.

(b)

Construct 90% confidence interval for ratio of two variances assembly

189

Chapter 4: Estimation

m2 en
times for men and women, 2
.
wom en
ANSWER EXERCISE CHAPTER 4
1.
(61.9616, 65.7184)

2.

(5552.120, 8530.735)

3.

(6.9186, 7.1914)

4.

(36.5104, 39.4895)

5.

(31897, 33023)

6.

(8.240, 8.362)

7.

(a)

0.12544

(b)

(0.990992, 1.349008)

(c)

(1.060496, 1.279504)

(d)

Yes.

8.

(0.1428, 4.1948)

9.

(0.9812, 1.1588)

10.

(-5.6104, 1.6104)

11.

(-9.7404, 13.5586)

12.

(-0.198, 1.998)

13.

(a)

(b)

(44.3625, 16.9966)

14.

(0.1757, 4.5764)

15.

(a)

(0.329, 0.870)

(b)

(1.253, 3.326)

(c)

Single-line system appears to be better. Because the variation appears

(7.2817, 24.8868)

to be significance lower with a single line.


16.

(a)

(0.308, 1.139)

(b)

17.

(a)

0.12544

(b)

(0.99092, 1.349008)

(c)

(0.99092, 1.349008)

18.

(a)

139.4242

(b)

(77.950, 335.228)

19.

(0.576, 2.213)

20.

(a)

(b)

(0.2398, 2.4663)

(0.2242, 1.0289)

190

(-0.33, 1.130)

Chapter 4: Estimation

SUMMARY CHAPTER 4
Confidence Interval for Single Mean
Z 2

Maximum error : E Z 2
, Sample size : n
n
E

(a)

(b)

Large Sample : n 30 or known

(i)

is known : x z / 2 / n x z / 2 / n

(ii)

is unknown : x z / 2 s / n x z / 2 s / n

Small Sample : n 30 and unknown

x t 2,v s / n x t 2,v s / n

; v n 1

Confidence Interval for a Difference Between Two Means


(a)

(b)

Z distribution case
(i)

2 2
is known : x1 x2 z / 2 1 2
n1
n2

(ii)

s2 s 2
is unknown : x1 x2 z / 2 1 2
n1
n2

t distribution case

1 2
2
s1 s 2
n

(i)

n1 n 2 , 12 22 :

x1 x2 t / 2,v

(ii)

n1 n 2 , 12 22 :

x1 x2 t / 2,v S p

S P2

(iii)

; v 2n 2

2
; v 2n 2

(n1 1) s12 (n2 1) s22


n1 n2 2

1
1
n1 n 2 , 12 22 : x1 x 2 t / 2,v S p
; v n1 n2 2
n1 n2

191

Chapter 4: Estimation

S P2

(n1 1) s12 (n2 1) s22


n1 n2 2
2

(iv)

s2 s 2
n1 n 2 , 12 22 : x1 x2 t / 2,v 1 2 ,
n1 n2

Confidence Interval for Single Population Variance

(n 1) s 2

2 / 2,v

(n 1) s 2

12 / 2,v

; v n 1

Confidence Interval for Ratio of Two Population Variances

1 s12
s12
1

f / 2,v 2,v1 ; v1 n1 1 and v2 n2 1


s22 f / 2,v1,v 2 22 s22
2

192

s12 s22

n n2
v 12
2
s12
s22


n1 n2
n1 1 n2 1

Chapter 4: Estimation

CORRECTION PAGE CHAPTER 4

193

Chapter 4: Estimation

194