Statistics For Management and Economics, 9th Edition Keller ISM PDF

How the Solutions Were Produced
All answers have been-double-checked for accuracy. However, we cannot be absolutely certain that
there are no errors. When and if we discover mistakes we will post corrected answers on our web
page. If you find any errors, please email the author (address on web page). We will be happy to
acknowledge you with the discovery.
Chapters 2 and 3
Excel was employed to draw the histograms, bar charts, pie charts, line charts, and scatter diagrams.
Chapter 4
Excel was used to draw box plots and compute the descriptive statistics for exercises with data sets.
Chapters 6 through 9
Probabilities were computed manually. Probability trees were used where possible.
Chapters 10 through 19 and 21
Calculations for exercises that provided statistics either in the exercise itself or in Appendix A were
completed manually. The solutions to exercises requiring the use of a computer were produced using
Excel. Confidence interval estimates used critical values obtained from the tables in Appendix B. In some
cases we were required to use approximations. As a consequence some confidence interval estimates will
differ slightly from those produced by computer. In tests of hypothesis where the sampling distribution is
normal, p-values were computed manually using Table 3. Excel was employed to calculate the p-value for
all other tests.
Chapters 13, and Appendixes 13 to 17, and 19
We employed the F-test of two variances at the 5% significance level to decide which one of the equal-
variances or unequal-variances t-test and estimator of the difference between two means to use to solve the
problem. Additionally, for exercises that compare two populations and are accompanied by data files, our
1
answers were derived by defining the sample from population 1 as the data stored in the first column (often
column A in Excel and column 1 in Minitab). The data stored in the second column represent the sample
from population 2. Paired differences were defined as the difference between the variable in the first
column minus the variable in the second column.
Chapter 19 and Appendix 19
In the exercises whose datasets contained interval data we used a nonparametric technique after examining
the relevant histograms and subjectively judging the variable to be “extremely nonnormal.”
Chapters 17 and 18
Excel produced all the solutions to these exercises.
Chapter 20
Most solutions were produced manually. Excel solved the more time-consuming exercises.
Chapter 21
All control charts were produced by Excel.
Chapter 22
Solutions to these exercises were completed manually.
2
TABLE OF CONTENTS
How the Solutions Were Produced 1
Chapter 1: What is Statistics? 3
Chapter 2: Graphical Descriptive Techniques I 5
Chapter 3: Graphical Descriptive Techniques II 37
Chapter 4: Numerical Descriptive Techniques 105
Chapter 5: Data Collection and Sampling 153
Chapter 6: Probability 157
Chapter 7: Random Variables and Discrete Probability Distributions 187
Chapter 8: Continuous Probability Distributions 215
Chapter 9: Sampling Distributions 229
Chapter 10: Introduction to Estimation 239
Chapter 11: Introduction to Hypothesis Testing 247
Chapter 12: Inference about One Population 269
Chapter 13: Inference about Two Populations 313
Appendix 13 Review of Chapters 12 and 13 375
Chapter 14: Analysis of Variance 393
Appendix 14 Review of chapters 12 to 14 483
Chapter 15: Chi-Squared Tests 457
Appendix 15 Review of Chapters 12 to 15 493
Chapter 16 Simple Linear Regression 509
Chapter 17: Multiple Regression 593
Chapter 18: Model Building 671
Chapter 19: Nonparametric Statistical Techniques 707
Chapter 20 Time-Series Analysis and Forecasting 793
Chapter 21 Statistical Process Control 815
Chapter 22: Decision Analysis 843
Instructor’s Solutions Manual
for
Statistics for Management and Economics
Ninth Edition
Prepared by
Gerald Keller
Wilfrid Laurier University

Chapter 1
1.2 Descriptive statistics summarizes a set of data. Inferential statistics makes inferences about
populations from samples.
1.3a The political choices of the 25,000 registered voters

b The political choices of the 200 voters interviewed
c Statistic
1.4a The complete production run

b 1000 chips
c Proportion of the production run that is defective
d Proportion of sample chips that are defective (7.5%)
e Parameter
f Statistic
g Because the sample proportion is less than 10%, we can conclude that the claim is true.
1.5 Survey graduates of your major as well as others and ask each person to report his or her
highest starting salary offer. Use statistical techniques to compare results.
1.6a Flip the coin 100 times and count the number of heads and tails
b Outcomes of flips
c Outcomes of the 100 flips
d Proportion of heads
e Proportion of heads in the 100 flips
1.8a The population consists of the fuel mileage of all the taxis in the fleet.
b The owner would like to know the mean mileage.
c The sample consists of the 50 observations.
d The statistic the owner would use is the mean of the 50 observations.
e The statistic would be used to estimate the parameter from which the owner can calculate total
costs.
We computed the sample mean to be 19.8 mpg.
3
4
Chapter 2
2.1 Nominal: Occupation, undergraduate major. Ordinal: Rating of university professor, Taste test
ratings. Interval: age, income
2.2 a Interval
b Interval
c Nominal
d Ordinal
2.3 a Interval
b Nominal
c Ordinal
d Interval
e Interval
2.4 a Nominal
b Interval
c Nominal
d Interval
e Ordinal
2.5 a Interval
b Interval
c Nominal
d Interval
e Nominal
2.6 a Interval
b Interval
c Nominal
d Ordinal
e Interval
2.7 a Interval
b Nominal
c. Nominal
5
d Interval
e Interval
f Ordinal
2.8 a Interval
b Ordinal
c Nominal
d Ordinal
2.9 a Interval
b Nominal
c Nominal
2.10 a Ordinal
b Ordinal
c Ordinal
2.11
300,000,000,000
250,000,000,000
200,000,000,000
150,000,000,000
100,000,000,000
50,000,000,000
0
6
2.12
United Venezuela, 8% Brazil, 1%

States, 2%
United A rab
Emirates, 8%
Canada, 14% China, 1%
Iran, 11%
Saudi A rabia,
22%
Iraq, 9 %
Kazakhstan, 2%
Russia, 5%
Kuwait, 8%
Qatar, 1 %
Nigeria, 3 % Libya, 4%
2.13
25,000,000
20,000,000
15,000,000
10,000,000
5,000,000
7
2.14
Liquified Lubricants Asphalt and Other

Residual fuel oil
refinery 1% road oil 2%
3%
gas 2%
Marketable 3%
coke
5%
Still gas
5%
Jet fuel Gasoline

13% 51%
Distillate fuel oil

15%
2.15
Hydroelectric Biomass Other

1.2% 4.1% 0.3%
Nuclear
0.0%
Natural gas
17.6% Coal
48.0%
Oil
28.8%
8
2.17
2.16
0
100
200
300
400
500
600
0
1000
2000
4000
6000
7000
3000
5000
Belgium
Brazil
Canada
China
France
Germany
India
Iran
Italy
Japan
9
Mexico
Poland
Russia
South Korea
Spain
Taiwan
Turkey
Ukraine
United Kingdom
United States
2.18
Yard trimmings, Other, 3%

13%
Paper and
paperboard,
34%
Food scraps,
12%
Wood, 5%
Textiles, 5% Glass, 5%

Rubber and
leather, 3 %Plastics, 1 2% Metals, 8%
2.19
Residential
Recycled Plastic, Recycled Glass,

1% 3%
Recycled
Disposed Other,
Recycled Paper, Metal, 1%
17%
14% Recycled
Organic/Food,
7%
Disposed Recycled
Organic, 23% Organic/ Yard,
10%
Discposed Recycled Other,

Paper, 12% 4%
Disposed Metal,
Disposed
2%
Plastic, 7 %
10
Non-Residential
Disposed O ther, Recycled Glass ,

Recycled
Disposed
6% 1% Metal, 3% Recycled
Construction/Dem
olition, 7% Organic, 1%
Recycled Recycled
Paper, 1 1% Construction/Dem
olition, 1%
Recycled
Other, 1 %
Disposed
Disposed Organic, Plastic, 1 0%
18%
Disposed Glass,
3%
Disposed Metal,
Disposed Paper,
8%
31%
2.20 a.
800
700
600
500
400
300
200
100
0
11
b.
Algeria
5% Angola
United Kingdom 6%
1% Venezuela
13%
Canada
Saudi A rabia 22%
17%
Nigeria
12% Iraq
Mexico
16% 6%
Kuwait
2%
The bar chart provides the frequencies and the pie chart displays the relative frequencies.
2.21
Males
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
Under 18 18-‐24 25-‐34 35-‐45 45-‐64
12
Females
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
Under 18 18-‐24 25-‐34 35-‐45 45-‐64
2.22
300
250
200
150
NFL
100 CFL
50
0
Four Parking Two ball Two beers Two drinks Four hot
tickets caps dogs
13
2.23
2.24
Pay medical Save Other

bills 5% 3%
6%
Buy something
40%
Invest
19%
Pay down debt

27%
14
2.25
Down payment Home

house Impulse improvement
2% 4% 13%
Appliances
Vacation 3%
11% Automobles
3%
Clothing
8%
Groceries Electronics
23% 8%
Gas Furniture
19% 3%
Spa/salon time
3%
2.26
Number of
Community
students
5%
8%
Career focus Location

16% 39%
Academic
reputation
10%
Majors
22%
15
2.27
Internet
8%
Word of mouth
12%
Consumer
guide
52%
Dealership
28%
2.28
Living/dining
room
9%
Kitchen Basement
27% 32%
Bedroom Bathroom
9% 23%
2.29 a Newspaper Frequency Relative Frequency

Daily News 141 .39
Post 128 .36
Times 32 .09
WSJ 59 .16
16
b
Wall Street
Journal
16%
New York Daily
news
New York Times 39%
9%
New York Post

36%
The Daily News and the Post dominate the market
2.30a Degree Frequency

BA 88
BBA 37
B Eng 51
B Sc 24
Other 30
b.
100
90
80
70
60
50
40
30
20
10
0
B.A. B.B.A. B.Eng B.Sc. Other
17
c
Other
13%
B.Sc. B.A.
11% 38%
B.Eng
22%
B.B.A.
16%
d. About 4 applicants in 10 have the BA degree, about one-fifth have a BEng. and one-sixth have a
BBA.
2.31a
45
40
35
30
25
20
15
10
5
0
IBM Compaq Dell Other
18
b
IBM
Other 21%
26%
Compaq
13%
Dell
40%
c Dell is most popular with 40% proportion, followed by other, 26%, IBM, 21% and Compaq,
13%.
2.32 a Software Frequency

Excel 34
Minitab 17
SAS 3
SPSS 4
Other 12
19
b
Other
17%
SPSS
6%
SAS Excel
4% 49%
Minitab
24%
c Excel is the choice of about half the sample, one-quarter have opted for Minitab, and a small
fraction chose SAS and SPSS.
2.33
Other
Natural Light 6%
9%
Bud L ight
31%
Miller L ite
21%
Busch L ight
Michelob L ight 7%
4%
Coors L ight
22%
20
2.34
Other
9%
Black
14%
White
77%
Three out of four Americans are White. Note that the survey did not separate Hispanics.
2.35
Graduate
degree Left high school
10% 15%
Bachelor's
degree
17%
Junior c ollege
8%
High school
50%
Half of American adults finished high school only. Approximately one-third received some kind
of post-secondary education.
21
2.36
Never married
24%
Married
Separated
48%
4%
Divorced
16%
Widowed
8%
Almost half the sample is married and about one out of four were never married.
2.37
Upper Lower
3% 7%
Middle
Working
47%
43%
Nine of out of 10 Americans classify themselves as middle or working class.
22
2.38
40,000
35,000
30,000
25,000
Less than high school
20,000 High school
15,000 Some c ollege

College graduate
10,000
5,000
0
1995 1999 2003 2007
The “Less than high school” category has remained constant, while the number of college
graduates has increased.
2.39
1200.0
1000.0
800.0
600.0
400.0
1990
200.0
1995
0.0
2000
2004
The state and local consumption has increased rapidlycompared to the other categories.
23
2.40 Australian Energy Sources
Hydroelectric Biomass Other

1.2% 4.1% 0.3%
Nuclear
0.0%
Natural gas
17.6% Coal
48.0%
Oil
28.8%
New Zealand Energy Sources
Other Coal
15.1% 7.0%
Biomass
4.4%
Oil
Hydroelectric 34.3%
10.1%
Nuclear
0.0%
Natural gas
29.1%
The dominant source in Australia is coal. In New Zealand it is oil.
24
2.41
400
350
300
250
200
1990
150
100 1995
50 2000
0 2006
Street crime has decreased while all the other categories have remained constant.
2.42
25
60
50
40
B.A.
30 B.Eng
B.B.A.
20
Other
10
0
University 1 University 2 University 3 University 4
Universities 1 and 2 are similar and quite dissimilar from universities 3 and 4, which also differ.
The two nominal variables appear to be related.
2.43
The column proportions are similar; the two nominal variables appear to be unrelated. There does
not appear to be any brand loyalty.
26
2.44
The two variables are related.
2.45
250
200
150
Men
100 Women
50
0
Lost job Left job Reentrants New e ntrants
There are large differences between men and women in terms of the reason for unemployment.
27
2.46 Counts
Column percent
50%
45%
40%
35%
30%
25%
20% Year: 1995
15% Year: 2000
10%
5% Year: 2007
0%
The number of prescriptions filled by independent drug stores has decreased while the others
remained constant or increased slightly.
28
2.47 Counts
Column percent
40%
35%
30%
25%
20%
15% Male
10%
Female
5%
0%
There appears to be differences between female and male students in their choice of light beer.
29
2.48
Excellent
Very good 3%
Poor
10% 15%
Good Fair
45% 27%
More than 40% rate the food as less than good.
2.49
Manual
18%
Computer
44%
Computer and
manual
38%
30
2.50 Canada
Other
11%
Yahoo
15%
Microsoft Google
12% 62%
United States
Other
18%
Google
36%
Yahoo
30%
Microsoft
16%
There are considerable differences between the two countries.
31
2.51
Strongly agree
4% Strongly
disagree
15%
Agree
23%
Disagree
20%
Neither agree
nor disagree
38%
More students disagree than agree.
2.52
32
45%
40%
35%
30%
25%
Children
20%
No c hildren
15%
10%
5%
0%
Poor Fair Good Very good Excellent
Customers with children rated the restaurant more highly than did customers with no children.
2.53
50.0
40.0
30.0
20.0
Manufacturers
10.0 Services
Aggregate
0.0
Belgium
Turkey
Sweden
Iceland
Australia
France
Korea
Denmark
Poland
Hungary
Canada
Spain
Germany
Luxembourg
Portugal
Austria
Greece
Mexico
Finland
Ireland
Netherlands
Norway
United States
New Zealand
Czech Republic
Slovak
Japan
Italy
United Kingdom
Switzerland
-‐10.0
There is a great deal of variation between countries. Moreover some countries tax manufacturers
more and some countries tax services more. The United States has the highest corporate tax rates.
33
2.54
45
40
35
30
25
20
15
10 Female
5
Male
0
Males and females differ in their areas of employment. Females tend to choose accounting
marketing/sales and males opt for finance.
34
40
35
30
25
20
15 Very satisfied
10 Quite satisfied
5
Little satisfied
0
Not satisfied
Area and job satisfaction are related. Graduates who work in finance and general management
appear to be more satisfied than those in accounting, marketing/sales, and others.
35
36
Chapter 3
3.1 9 or 10
3.2 10 or 11
3.3 a 7 to 9
188 − 37
b Interval width ≈ = 18.9 (rounded to 20); upper limits: 40, 60, 80,100,120, 140, 160 180,
8
200
3.4 a 7 to 9
6.1 − 5.2
b Interval width ≈ = .13 (rounded to .15); upper limits: 5.25, 5.40, 5.55, 5.70, 5.85, 6.00,
7
6.15
3.5a
15
Frequency
10
0
5 10 15 20 25
Defectives
1.00
0.80
0.60
0.40
0.20
0.00
0 5 10 15 20 25
Defectives
c The histogram is unimodal and somewhat positively skewed.
37
3.6 a
10
8
Frequency 6
4
2
0
4 6 8 10 12
Pages
1.00
0.80
0.60
0.40
0.20
0.00
0 2 4 6 8 10 12
Pages
c The number of pages is bimodal and slightly positively skewed.
3.7 a
15
Frequency
10
0
30 36 42 48 54 60 66
Ages
38
c
1.00
0.80
0.60
0.40
0.20
0.00
24 30 36 42 48 54 60 66 More
Ages
d The ages are bimodal and positively skewed.
3.8
15
Frequency
10
0
5 10 15 20 25
Calls
The histogram is bimodal.
3.9 a
Stem & Leaf Display
Stems Leaves
30 ->0112222222356667777788
31 ->001113568
32 ->024777
33 ->0047
34 ->024455
35 ->7
36 ->7
37 ->9
39
b
30
Frequency
20
10
0
31 32 33 34 35 36 37 38 39 More
Times
c The histogram is positively skewed.
3.10 a
25
20
Frequency
15
10
5
0
1 2.5 4 5.5 7 8.5 10 More
Stores
1.00
0.80
0.60
0.40
0.20
0.00
0 1 2.5 4 5.5 7 8.5 10
Stores
c The number of stores is bimodal and positively skewed.
40
3.11
20
15
Frequency 10
5
0
8 12 16 20 24 28
Games
The histogram is positively skewed.
3.12 a
20
15
Frequency
10
5
0
5 10 15 20 25 30 35 40 45
Rounds
1.00
0.80
0.60
0.40
0.20
0.00
0 5 10 15 20 25 30 35 40 45
Rounds
41
d The histogram is symmetric (approximately) and bimodal.
3.13
80
Frequency
60
40
20
0
The histogram is symmetric, unimodal, and bell shaped.
3.14 a
25
20
Frequency
15
10
5
0
500 650 800 950 1100 1250 1400
Prices
1.00
0.80
0.60
0.40
0.20
0.00
350 500 650 800 950 1100 1250 1400
Prices
42
c
A B C
1 Stem & Leaf Display
2
3 Stems Leaves
4 4 ->2445677789
5 5 ->0122224668899
6 6 ->0001244555667
7 7 ->00022237889
8 8 ->01333445667
9 9 ->012246667788
10 10 ->00233788
11 11 ->015
12 12 ->18
13 13 ->23
d The histogram is slightly positively skewed, unimodal, and not bell-shaped.
3.15
80
60
Frequency
40
20
0
25 30 35 40 45 50 55 60 65 70
Customers
The histogram is unimodal and positively skewed.
3.16 a The histogram should contain 9 or 10 bins. We chose 10.

b
100
80
Frequency
60
40
20
0
40 46 52 58 64 70 76 82 88 94
Times
c The histogram is positively skewed.

d The histogram is not bell-shaped.
43
3.17
100
80
Frequency 60
40
20
0
37 44 51 58 65 72 79 86 93 100
Marks
The histogram is negatively skewed, bimodal, and not bell shaped.
3.18
60
Frequency
40
20
0
18 19 20 21 22 23 24
Lengths
The histogram is unimodal, bell-shaped and roughly symmetric. Most of the lengths lie between
18 and 23 inches.
3.19
30
Frequency
20
10
0
200 400 600 800 1000 1200 1400
Copies
The histogram is unimodal and positively skewed. On most days the number of copies made is
between 200 and 1000. On a small percentage of days more than 1000 copies are made.
44
3.20
100
80
Frequency 60
40
20
0
1 2 3 4 5 6 7 8
Weights
The histogram is unimodal, symmetric and bell-shaped. Most tomatoes weigh between 2 and 7
ounces with a small fraction weighing less than 2 ounces or more than 7 ounces.
3.21
150
Frequency
100
50
0
15 20 25 30 35 40 45 50 55
Gallons
The histogram is positively skewed and unimodal. Most households use between 20 and 45
gallons per day. The center of the distributions appears to be around 25 to 30 gallons.
3.22
30
Frequency
20
10
0
12000 15000 18000 21000 24000 27000 30000
Books
The histogram of the number of books shipped daily is negatively skewed. It appears that there is
a maximum number that the company can ship.
45
3.23 a
80
60
Frequency 40
20
0
450 500 550 600 650 700 750 800 850 900
Repaid
b
30
Frequency
20
10
0
450 500 550 600 650 700 750 800 850 900
Defaulted
c The scorecards appear to be relatively poor predictors.
3.24 a
100
80
Frequency
60
40
20
0
450 500 550 600 650 700 750 800 850 900
Repaid
46
b
30
Frequency
20
10
0
450 500 550 600 650 700 750 800 850 900
Defaulted
c. and d. This scorecard is a much better predictor.
3.25a Interval
b
800
600
Frequency
400
200
0
0 1 2 3 4 5 6 7 8 9 1011121314151617181920
EDUC
c The peaks in the histogram represent high school graduates , two-year college graduates, and
university graduates.
3.26
400
300
Frequency
200
100
0
1 2 3 4 5 6 7 8 More
TVHOURS
The histogram is highly positively skewed indicating that most people watch 4 or less hours per
day with some watching considerably more.
47
3.27
250
200
Frequency
150
100
50
0
20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
AGE
3.28
500
400
Frequency
300
200
100
0
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
HRS
Many people work more than 40 hours per week.
3.29
2000
1800
1600
1400
1200
1000
800
600
400
200
0
1 2 3 4 5 6 7 8 9 10 11 12 13
48
3.30
80
70
60
50
40
30
20
10
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
After a rapid increase the numbers have leveled off.
3.31
2500
2000
1500
1000
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Total health care expenditures are rising faster than inflation
49
3.32
2000
1800 Female
1600
1400
Male
1200
1000
800
600
400
200
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
The numbers of females and males are both increasing with the number of females increasing
faster.
3.33
14000
12000
10000
8000
6000
4000
2000
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
The number of property crime decreased slowly over the 15 years.
50
3.34
6000
5000
4000
3000
2000
1000
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
The per capita number of property crimes decreased faster than did the absolute number of
property crimes.
3.35a
16000
14000
12000
10000
8000
6000
4000
2000
0
1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829
GDP increased rapidly over the 29 year period.
51
b
8000
7000
6000
5000
4000
3000
2000
1000
0
1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829
The inflation-adjusted GDP grew at a moderate rate.
3.36
25,000
20,000
Consumption
15,000
10,000
5,000 Production
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
Consumption is increasing and production is falling.
52
3.37
200,000
180,000
160,000
140,000
All
120,000
Northeast
100,000
Midwest
80,000
South
60,000
40,000 West
20,000
-‐
1 2 3 4 5 6 7 8 9 1011121314151617181920
All areas as well as the whole country saw house prices staying ahead of inflation until the last
three years.
3.38
3000
2500
2000
1500 Receipts
Outlays
1000
500
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27
53
b
300
200
100
0
1 2 3 4 5 6 7 8 9 10111213141516171819202122232425262728
-‐100
-‐200
-‐300
-‐400
-‐500
Over the last 28 years both receipts and outlays increased rapidly. There was a five-year period
where receipts were higher than outlays. Between 2004 and 2007 the deficit has decreased.
3.39
0.03
0.02
0.01
0
-‐0.01 1 2 3 4 5 6 7 8 9 10111213141516171819202122232425262728
-‐0.02
-‐0.03
-‐0.04
-‐0.05
-‐0.06
-‐0.07
When the size of the economy as measured by GDP the deficits are not that large.
54
3.40
0.02
0.01
0
1 2 3 4 5 6 7 8 9 10111213141516171819202122232425262728
-‐0.01
-‐0.02
-‐0.03
-‐0.04
-‐0.05
-‐0.06
-‐0.07
The inflation adjusted deficits are not large.
3.41 The cost is calculated as follows.

100 Dis tan ce ×1,000
Cost per year in 1982-84 dollars = Pr ice ×
CPI MPG
1200
1000
800
600
400
200
0
1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829
Even though the average distance travelled per year has increased the annual inflation-adjusted
cost of driving decreased from over $1,000 in 1980 to less than $600 in 2005 before starting to
increase from 2002 to 2006.
55
3.42 Exports to Canada
30000
25000
20000
15000
10000
5000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
Imports from Canada
35000
30000
25000
20000
15000
10000
5000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
56
Balance of trade: Exports to Canada – Imports from Canada
0
1 25 49 73 97 121 145 169 193 217 241 265 289
-‐2000
-‐4000
-‐6000
-‐8000
-‐10000
-‐12000
Imports from Canada have greatly exceeded exports to Canada.
3.43 Exports to Japan
7000
6000
5000
4000
3000
2000
1000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
57
Imports from Japan
16000
14000
12000
10000
8000
6000
4000
2000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
Balance of trade: Exports to Japan – Imports from Japan
0
1 25 49 73 97 121 145 169 193 217 241 265 289
-‐1000
-‐2000
-‐3000
-‐4000
-‐5000
-‐6000
-‐7000
-‐8000
-‐9000
Imports from Japan have greatly exceeded exports to Japan.
58
3.44
Canadian Dollars to One U.S. Dollar

1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
1 25 49 73 97 121145169193217241265289313337361385409433457
In the early seventies the Canadian dollar was worth more than the U.S. dollar. By the late
seventies the Canadian lost ground but has recently recovered.
3.45
400
350
300
250
200
150
100
50
0
1 25 49 73 97 121145169193217241265289313337361385409433457
The yen has significantly increased its value in terms of the American dollar.
59
3.46
16000
14000
12000
10000
8000
6000
4000
2000
0
1 61 121 181 241 301 361 421 481 541 601 661
The index grew slowly until month 400 and then grew quickly until month 600. It then fell sharply
and recently recovered.
3.47
8000
7000
6000
5000
4000
3000
2000
1000
0
1 61 121 181 241 301 361 421 481 541 601 661
The inflation-adjusted index displays far less volatility.
60
3.48
1.65
Movie attendnace 1.6
1.55
1.5
1.45
1.4
1.35
0 10 20 30 40 50 60 70 80
DVD percentage
There does not appear to be a linear relationship between the two variables.
3.49a
40
30
20
Returns
10
0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
-‐10
-‐20
Inflation
b. There is very weak linear relationship.
61
3.50a
90
85
80
Statistics
75
70
65
60
40 50 60 70 80 90 100
Calculus
b. There is a positive linear relationship between calculus and statistics marks.
3.51
2500
2000
1500
Cost
1000
500
0
0 5 10 15 20 25
Speed
There is a strong positive linear relationship.
62
3.52a
30
25
20
Internet use
15
10
0
0 5 10 15 20
Education
b. There is a moderately strong positive linear relationship. In general those with more education
use the Internet more frequently.
3.53
100
95
90
85
80
Mark
75
70
65
60
55
50
60 65 70 75 80 85 90 95
Time
There is no linear relationship
63
3.54a
1600
1400
1200
1000
Electricity
800
600
400
200
0
0 1 2 3 4 5 6 7 8 9
Occupants
b. There is a moderately strong positive linear relationship.
3.55
700
600
500
400
Tickets
300
200
100
0
-‐40 -‐30 -‐20 -‐10 0 10 20 30 40
Temperature
There is a moderately strong linear relationship. More lift tickets are sold during warmer days.
64
3.56a
90
80
70
60
50
Income
40
30
20
10
0
60 65 70 75 80
Height
b. There is a very weak positive linear relationship.
3.57
2000000
1500000
Compensation
1000000
500000
0
0 5000 10000 15000 20000 25000 30000
-‐500000
Profit
65
500
450
400
Compensation 350
300
250
200
150
100
50
0
0 5000 10000 15000 20000 25000 30000
3-‐Yr share return
There does not appear to be a linear relationship between compensation of profit and between
compensation and 3-year share return.
3.58
600
500
400
Tenure
300
200
100
0
0 20 40 60 80 100
Age
There is a moderately strong positive linear relationship.
66
3.59
3000
2500
2000
Sales
1500
1000
500
0
0 5 10 15 20 25 30 35
Time
There is no linear relationship.
3.60
400
350
300
Livestock sub-‐index
250
200
150
100
50
0
0 100 200 300 400 500 600
Grains sub-‐index
There is moderately strong positive linear relationship.
67
3.61
25
Bank prime loan rate 20
15
10
0
0 2 4 6 8 10 12
Unemployment rate
There is a weak positive linear relationship.
3.62
18
16
14
12
Education
10
8
6
4
2
0
0 20 40 60 80 100
Age
There does not appear to be any relationship between the two variables.
68
3.63
1000
900
800
700
600
TIME2
500
400
300
200
100
0
0 100 200 300 400 500 600
TIME1
There is a weak positive linear relationship between the two variables.
3.64
1000
900
800
700
600
TIME3
500
400
300
200
100
0
0 100 200 300 400 500 600
TIME1
There does not appear to be a linear relationship.
3.65 There is one person who claimed to spend 600 minutes on an average day reading a printed
newspaper. This is probably an error. We deleted that number and drew a new scatter diagram.
There now appears to be a weak positive linear relationship.
69
200
180
160
140
120
TIME3
100
80
60
40
20
0
0 100 200 300 400 500 600
TIME1
3.66
600
500
400
TIME4
300
200
100
0
0 200 400 600 800 1000 1200 1400
TIME2
There does not appear to be a linear relationship between the two variables.
70
3.67
600
500
400
TIME1
300
200
100
0
0 20 40 60 80 100
AGE
There does not appear to be any linear relationship between age and the amount of time spent
reading news on the Internet.
3.68
25
20
15
SPEDUC
10
0
0 5 10 15 20 25
EDUC
There is a moderately strong positive linear relationship between the education levels of spouses.
71
3.69
25
20
15
EDUC
10
0
0 5 10 15 20 25
PAEDUC
There is a weak positive linear relationship between the amount of education of fathers and their
children.
3.70
25
20
15
EDUC
10
0
0 5 10 15 20 25
MAEDUC
There is a weak positive linear relationship between the amount of education of mothers and their
children.
72
3.71
100
90
80
70
60
SPHRS
50
40
30
20
10
0
0 20 40 60 80 100
HRS
There is no linear relationship between the amounts of hours of work per week between spouses.
3.72 Region: Sales last year
Australasia
17%
United States
44%
Europe
25%
Canada
14%
73
Region: Sales previous year
Austral
asia
10% United
States
38%
Europe
34%
Canada
18%
The pie charts were drawn so that the area in each pie is proportion to the total sales in each year.
For example to draw the first pie chart we solved for the radius as follows.
Πr2 = 152.3
Solving for r we find
r = 6.96
For the second pie chart we find r = 5.78.
We draw the pie charts for the divisions in each year in the same way.
Division: Sales last year
Financial Hotels and

systems clubs
10% 3%
Manufacturing
9%
Customer
service
Construction 42%
7%
Library systems
29%
74
Division: Sales previous year
Financial Hotels and

systems clubs
Manufactur
6% 4%
ing
10%
Constructio
Customer
n
service
12%
36%
Library
systems
32%
3.73a
14,000
12,000
10,000
8,000
Violent c rimes
6,000
Property c rimes
4,000
2,000
0
1 3 5 7 9 11 13 15 17 19 21
b. It appears that the number of crimes have stayed the same of decreased slightly.
c. We need other variables to try to explain these data.
3.74 We divided the number of crimes by the population and multiplied by 1,000. The result is the
number of crimes per thousand of population.
75
60
50
40
30 Series1
Series2
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
b We can see that there has been a decrease in the number of crimes per thousand of population
c Another possible chart is a scatter diagram of the number of crimes and population.
Violent crimes
2,500
2,000
Violent crimes
1,500
1,000
500
0
200,000 220,000 240,000 260,000 280,000 300,000 320,000
Population (1,000)
76
Property crimes
14,000
12,000
10,000
Property crimes
8,000
6,000
4,000
2,000
0
200,000 220,000 240,000 260,000 280,000 300,000 320,000
Population (1,000)
The unusual shapes of the scatter diagram are difficult to explain. Both charts are quadratic rather
than linear. They suggest that when the population was less than 255 million the population and
the number of crimes were positively linear related, whereas when the population was more than
255 million the relationship was negative.
3.75 Many experts believe that a large proportion of the violent and property crimes are committed
by people age 15 to 24.
2,500
2,000
Violent crimes
1,500
1,000
500
0
35,000 36,000 37,000 38,000 39,000 40,000 41,000 42,000 43,000
Population:15-‐24(1,000)
77
Scatter diagram of number of property crimes and population 15 to 24
14,000
12,000
10,000
Property crimes
8,000
6,000
4,000
2,000
0
35,000 36,000 37,000 38,000 39,000 40,000 41,000 42,000 43,000
Population:15-‐24(1,000)
The scatter diagrams indicate that there is a negative linear relationship between the number of
crimes and the population aged 15 to 24, refuting the experts.
3.76 a We convert the numbers to accident rate and fatal accident rate.
A B C
1 Age group Accident rate per driver Fatal accidient rate (per 1,000 drivers)
2 Under 20 0.373 0.643
3 20-24 0.173 0.352
4 25-34 0.209 0.305
5 35-44 0.162 0.251
6 45-54 0.133 0.214
7 55-64 0.108 0.208
8 65-74 0.095 0.177
9 0ver 74 0.093 0.304
78
b
0.7
0.6
0.5
0.4
Accidents/driver
0.3
Fatal accidents/1000
0.2 drivers
0.1
0
Under 20-‐24 25-‐34 35-‐44 45-‐54 55-‐64 65-‐74 0ver
20 74
c. The accident rate generally decreases as the ages increase. The fatal accident rate decreases until
the over 64 age category where there is an increase.
3.77
79
2.500
2.000
1.500
1.000 Injury r ate

Fatal injury r ate per 100
0.500
0.000
c Older drivers who are in accidents are more likely to be killed or injured.
d Exercise 3.9 addressed the issue of accident rates, whereas in this exercise we consider the
severity of the accidents.
3.78a
600
550
500
450
400 Verbal A ll
350 Math A ll
300
250
200
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
80
b
550
540
530
520 Verbal A ll
Math A ll
510
500
490
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
560
540
520
500
480 Verbal Male
460 Verbal Female
440
420
400
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
81
600
500
400
300 Math Male

Math Female
200
100
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
550
540
530
520 Verbal Male

Verbal Female
510
500
490
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
82
550
540
530
520
510
Math Male
500
Math Female
490
480
470
460
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
3.79a
7.8
7.6
7.4
7.2
7.0
6.8
6.6
6.4
6.2
6.0
1 2 3 4 5 6 7 8 9 10 11 12
83
b
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1 2 3 4 5 6 7 8 9 10 11 12
c Caption a: Unemployment rate falling rapidly. Caption b: Unemployment rate virtually

unchanged.
d The chart in (a) is more honest.
3.80 a
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53
b To show actual changes it is probably best to show constant dollars on a graph with a 0.
84
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53
3.81 a Private Schools

K-Grade 8 Enrollment: All years
5,100
4,900
4,700
4,500
4,300
4,100
3,900
3,700
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
85
K-Grade 8 Enrollment: Year 15 - Year 37 (1979 – 2001)
5,200
5,000
4,800
4,600
4,400
4,200
4,000
3,800
3,600
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Grade 9 – Grade 12 Enrollment: All years
1,450
1,400
1,350
1,300
1,250
1,200
1,150
1,100
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Grade 9 – Grade 12 Enrollment: Years 29-41 (1993-2005)
86
1,400
1,350
1,300
1,250
1,200
1,150
1,100
1 2 3 4 5 6 7 8 9 10 11 12 13
College Enrollment: All years
4,900
4,400
3,900
3,400
2,900
2,400
1,900
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
87
b Public Schools
K-Grade 8 Enrollment: All years
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Grade 9 – Grade 12 Enrollment: All years
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
88
College Enrollment
50,000
45,000
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
3.82 a
Married
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
89
b
Single
69.0
67.0
65.0
63.0
61.0
59.0
57.0
55.0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
3.83
40%
30%
20%
Return on sub-‐index
10%
0%
-‐10%
-‐20%
-‐30%
-‐40%
-‐1.00% -‐0.50% 0.00% 0.50% 1.00% 1.50% 2.00%
Inflation rate
There is no linear relationship between the inflation rate and the return on the precious metals sub-
index.
90
3.84
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
1 61 121 181 241 301 361 421
There has been a long-term decline in the value of the Australian dollar.
3.85
160
140
120
100
IQ Twin 2
80
60
40
20
0
0 20 40 60 80 100 120 140
IQ Twin 1
There is a strong positive linear relationship.
91
3.86
180
160
140
Currencies index
120
100
80
60
40
20
0
60 70 80 90 100 110 120 130 140
Interest rate index
There is a very strong positive linear relationship.
3.87
30
Frequency
20
10
0
10 15 20 25 30 35 40 45
Crashes
The histogram of the number of crashes is positively skewed.
92
3.88a
74
72
70
Son's height
68
66
64
62
60
62 64 66 68 70 72 74 76
Father's height
b. The slope is positive

c. There is a moderately strong linear relationship.
3.89
12
10
Unemployment rate
0
0 2000 4000 6000 8000 10000 12000 14000 16000
DJIA
There is no linear relationship between the Dow Jones Industrial average and the unemployment
rate.
93
3.90
2.4
2.2
1.8
1.6
1.4
1.2
1
1 25 49 73 97 121 145 169 193 217 241 265
The value of the British pound has fluctuated quite a bit but the current exchange rate is close to
the value in 1987.
3.91
350
330
310
290
270
Time
250
230
210
190
170
150
300 320 340 360 380 400 420
Score
There is a strong positive linear relationship. Poorer players take longer to complete their rounds.
94
3.92a
Imports from Mexico
25000
20000
15000
10000
5000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
b. Exports to Mexico
16000
14000
12000
10000
8000
6000
4000
2000
0
1 25 49 73 97 121 145 169 193 217 241 265 289
95
c. Balance of Trade (Imports from Mexico – exports to Mexico)
8000
7000
6000
5000
4000
3000
2000
1000
0
-‐1000 1 25 49 73 97 121 145 169 193 217 241 265 289
-‐2000
d. The United States imports more products from Mexico than it exports to Mexico. Moreover, the
trade imbalance is worsening (only interrupted by the recession in 2008-2009).
3.93
1400
1200
1000
800
Credit
600
400
200
0
0 100 200 300 400 500
Debit
There is a moderately strong negative linear relationship.
96
3.94
40
30
Frequency 20
10
0
25000 40000 55000 70000 85000 100000
Pay
3.95
50
40
Frequency
30
20
10
0
7 10 13 16 19 22 25
Meetings
3.96
Fatal accidents
35
30
25
20
15
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
97
Passenger deaths
1400
1200
1000
800
600
400
200
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
The number of fatal accidents and the number of deaths have been decreasing.
3.97
200
150
Frequency
100
50
0
10 20 30 40 50 60 70 80 90 100 110
Miles
3.98
200
150
Frequency
100
50
0
30 45 60 75 90 105 120 135 150 165
Times
The histogram tells us that about 70% of gallery visitors stay for 60 minutes or less and most of
the remainder leave within 120 minutes. Although there are other plans, the gallery director
98
proposed the following plan. Admit 200 visitors every hour. We expect that about 140 will leave
within 1 hour and about 60 will stay for an additional hour. During the next 1-hour period, 200
new visitors will be admitted. If 60 of the previous hour’s admittances remain, there will be a total
of 260 people in the gallery. If this pattern persists during the day, there will be a maximum of 260
visitors at any time. This plan should permit as many people as possible to see the exhibit and yet
maintain comfort and safety.
3.99 Business Statistics course (Example 3.3)
100
90
80
Statistics
70
60
50
40
25 50 75 100
Calculus
Mathematical Statistics course (Example 3.4)
100
90
80
Statistics
70
60
50
40
30
25 50 75 100
Calculus
99
There appears to be a stronger linear relationship between marks in the mathematical statistics
course and calculus than the relationship between the marks in the business statistics course and
the marks in calculus.
3.100 Business Statistics course (Example 3.3)
100
90
80
Final
70
60
50
40
50 60 70 80 90 100
Midterm
Mathematical Statistics course (Example 3.4)
100
90
80
70
Final
60
50
40
30
50 60 70 80 90 100
Midterm
The relationship between midterm marks and final marks appear to be similar for both statistics
courses. That is, there is weak positive linear relationship.
100
3.101a
80
60
Frequency 40
20
0
30 34 38 42 46 50 54 58 62 66
Salary
The histogram is approximately bell shaped and symmetric.
70
60
50
40
Salary
30
20
10
0
0 1 2 3 4 5 6 7
Search
There is no linear relationship between the amount of time needed to land a job and salary.
101
Case 3.1 Line Chart of Temperature Anomalies
2.5
2
1.5
1
0.5
0
-‐0.5
-‐1
-‐1.5
-‐2
-‐2.5
1 121 241 361 481 601 721 841 961 1081 1201 1321 1441
There is a clear upward trend of about 1 degree Celsius over the 130 years.
Scatter Diagram of National Climate Data Center Land and Sea Temperature Anomalies and CO2
Levels (1958-2009)
1.5
Temperature anomalies
0.5
-‐0.5
-‐1
-‐1.5
300 320 340 360 380 400
CO2
There is a moderately strong positive linear relationship between carbon dioxide levels and
temperature anomalies.
102
Case 3.2
90,000
80,000
70,000
60,000
50,000
GDP
40,000
30,000
20,000
10,000
0
0 20 40 60 80 100
Freedom scores
There is a moderately strong positive linear relationship between freedom scores and gross
domestic product.
103
104
Chapter 4
4.1 a x =
∑x i
=
52 + 25 + 15 + 0 + 104 + 44 + 60 + 30 + 33 + 81 + 40 + 5 489
= = 40.75
n 12 12
Ordered data: 0, 5, 15, 25, 30, 33, 40, 44, 52, 60, 81, 104; Median = (33 + 40)/2 = 36.5
Mode = all
4.2 x =
∑x i
=
5 + 7 + 0 + 3 + 15 + 6 + 5 + 9 + 3 + 8 + 10 + 5 + 2 + 0 + 12 90
= = 6.0
n 15 15
Ordered data: 0, 0, 2, 3, 3, 5, 5, 5, 6, 7, 8, 9, 10, 12, 15; Median = 5
Mode = 5
4.3 a x =
∑x i
=
5.5 + 7.2 + 1.6 + 22. 0 + 8.7 + 2.8 + 5.3 + 3.4 + 12.5 + 18.6 + 8.3 + 6.6 102.5
= =
n 12 12
8.54
Ordered data: 1.6, 2.8, 3.4, 5.3, 5.5, 6.6, 7.2, 8.3, 8.7, 12.5, 18.6, 22.0; Median = 6.9
Mode = all
b The mean number of miles jogged is 8.54. Half the sample jogged more than 6.9 miles and half
jogged less.
4.4 a x =
∑x i
=
33 + 29 + 45 + 60 + 42 + 19 + 52 + 38 + 36 354
= = 39.3
n 9 9
Ordered data: 19, 29, 33, 36, 38, 42, 45, 52, 60; Median = 38
Mode: all
b The mean amount of time is 39.3 minutes. Half the group took less than 38 minutes.
4.5 a x =
∑x i
=
14 + 8 + 3 + 2 + 6 + 4 + 9 + 13 + 10 + 12 + 7 + 4 + 9 + 13 + 15 + 8 + 11 + 12 + 4 + 0
n 20
164
= = 8.2
20
Ordered data: 0, 2, 3, 4, 4, 4, 6, 7, 8, 8, 9, 9, 10, 11, 12, 12, 13, 13, 14, 15; Median = 8.5
Mode = 4
b The mean number of days to submit grades is 8.2, the median is 8.5, and the mode is 4.
105
4.6 R g = 3 (1 + R1)(1 + R 2 )(1 + R 3 ) − 1 = 3 (1 + .25)(1 − .10)(1 + .50) − 1= .19
4.7 R g = 4 (1 + R1)(1 + R 2 )(1 + R 3 )(1 + R 4 ) − 1 = 4 (1 + .50)(1 + .30)(1 − .50)(1 − .25) − 1 = –.075
4.8 a x =
∑x i
=
.10 + .22 + .06 − .05 + .20 .53
= = .106
n 5 5
Ordered data: –.05, .06, .10, .20, .22; Median = .10
b R g = 5 (1 + R1)(1 + R 2 )(1 + R 3 )(1 + R 4 )(1 + R 5 ) − 1 =
5 (1 + .10)(1 + .22)(1 + .06)(1 − .05)(1 + .20) − 1 = .102
c The geometric mean is best.
4.9 a x =
∑x i
=
- .15 - .20 + .15 − .08 + .50 .22
= = .044
n 5 5
Ordered data: –.20, –.15, –.08, .15, .50; Median = –.08
b R g = 5 (1 + R1)(1 + R 2 )(1 + R 3 )(1 + R 4 )(1 + R 5 ) − 1 =
5 (1 − .15)(1 − .20)(1 + .15)(1 − .08)(1 + .50) − 1= .015
c The geometric mean is best.
1200 − 1000
4.10 a Year 1 rate of return = = .20
1000
1200 − 1200
Year 2 rate of return = =0
1200
1500 − 1200
Year 3 rate of return = = .25
1200
2000 − 1500
1500
b x=
∑x i
=
.20 + 0 + .25 + .33 .78
= = .195
n 4 4
Ordered data: 0, .20, .25, .33; Median = .225
c R g = 4 (1 + R1)(1 + R 2 )(1 + R 3 )(1 + R 4 ) − 1 = 4 (1 + .20)(1 + 0)(1 + .25)(1 + .33) − 1 = .188
d The geometric mean is best because 1000(1.188) 4 = 2000.
106
10 − 12
4.11 a Year 1 rate of return = = –.167
12
14 − 10
10
15 − 14
14
22 − 15
15
30 − 22
22
25 − 30
Year 6 rate of return = = –.167
30
b x=
∑x i
=
- .167 + .40 + .071 + .467 + .364 − .167 .968
= = .161
n 6 6
Ordered data: –.167, –.167, .071, .364, .40, .467; Median = .218
c R g = 6 (1 + R 1 )(1 + R 2 )(1 + R 3 )(1 + R 4 )(1 + R 5 )(1 + R 6 ) − 1
= 6 (1 − .167)(1 + .40)(1 + .071)(1 + .467)(1 + .364)(1 − .167) − 1 = .130
d The geometric mean is best because 12(1.130)6 = 25.
4.12 a x = 75,750; median = 76,410

b The mean starting salary is $75,750. Half the sample earned less than $76,410.
4.13 a x = 11.19; median = 11

b The mean number of days is 11.19 and half the sample took less than 11 days and half took more
than 11 days to pay.
4.14a x = 117.08; median = 124.00

b The mean expenditure is $117.08 and half the sample spent less than $1246.00.
4.15a x = 26.80; median = 27.00

b x = 30.94; median = 31.00
c The mean and median of commuting time in New York is larger than that in Los Angeles.
4.16a x = .81; median = .83

b The mean percentage is .81. Half the sample paid less than .83.
4.17a x = 32.91; median = 32; mode = 32
107
b The mean speed is 32.91 mph. Half the sample traveled slower than 32 mph and half traveled
faster. The mode is 32.
4.18a x = 592.04; median = 591.00

b The mean expenditure is $592.04. Half the sample spent less than $591.00
4.19 x =
∑x i
=
9 + 3 + 7 + 4 + 1 + 7 + 5 + 4 40
= =5
n 8 8
2 ∑ (x i − x) 2 [(9 − 5) 2 + (3 − 5) 2 + ... + (4 − 5) 2 46
s = = = = 6.57
n −1 8 −1 7
4.20 x =
∑x i
=
4 + 5 + 3 + 6 + 5 + 6 + 5 + 6 40
= =5
n 8 8
∑ (x i − x) 2 [(4 − 5) 2 + (5 − 5) 2 + ... + (6 − 5) 2 8
s2 = = = = 1.14
n −1 8 −1 7
4.21 x =
∑x i
=
12 + 6 + 22 + 31 + 23 + 13 + 15 + 17 + 21 160
= = 17.78
n 9 9
∑ (x i − x) 2 [(12 − 17.78) 2 + (6 − 17.78) 2 + ... + (21 − 17.78) 2 433.56

s2 = = = = 54.19
n −1 9 −1 8
s = s 2 = 54.19 = 7.36
4.22 x =
∑x i
=
0 + (−5) + (−3) + 6 + 4 + (−4) + 1 + (−5) + 0 + 3
=
−3
= –.30
n 10 10
∑ (x i − x) 2 [(0 − (−.3)) 2 + ((−5) − (−.3)) 2 + ... + (3 − (−.3)) 2 136.1

s2 = = = = 15.12
n −1 10 − 1 9
s = s 2 = 15.12 = 3.89
4.23 The data in (b) appear to be most similar to one another.
4.24 a: s 2 = 51.5
108
b: s 2 = 6.5
c: s 2 = 174.5
4.25 Variance cannot be negative because it is the sum of squared differences.
4.26 6, 6, 6, 6, 6
4.27 a about 68%

b about 95%
c About 99.7%
4.28 a From the empirical rule we know that approximately 68% of the observations fall between
46 and 54. Thus 16% are less than 46 (the other 16% are above 54).
b Approximately 95% of the observations are between 42 and 58. Thus, only 2.5% are above 58
and all the rest, 97.5% are below 58.
c See (a) above; 16% are above 54.
4.29 a at least 75%

b at least 88.9%
4.30 a Nothing
b At least 75% lie between 60 and 180.
c At least 88.9% lie between 30 and 210.
4.31 Range = 25.85, s 2 = 29.46, and s = 5.43; there is considerable variation between prices; at
least 75% of the prices lie within 10.86 of the mean; at least 88.9% of the prices lie within 16.29
of the mean.
4.32 s 2 = 40.73 mph 2 and s = 6.38 mph; at least 75% of the speeds lie within 12.76 mph of the
mean;
at least 88.9% of the speeds lie within 19.14 mph of the mean
4.33 a Punter Variance Standard deviation

1 40.22 6.34
2 14.81 3.85
3 3.63 1.91
b Punter 3 is the most consistent.
109
4.34 s 2 = .0858 cm2, and s = .2929 cm; at least 75% of the lengths lie within .5858 of the mean; at
least 88.9% of the rods will lie within .8787 cm of the mean.
4.35 x = 175.73 and s = 62.1; At least 75% of the withdrawals lie within $124.20 of the mean; at
least 88.9% of the withdrawals lie within $186.30 of the mean..
4.36a s = 15.01
b In approximately 68% of the days the number of arrivals falls within 15.01 of the mean; in
approximately 95% of the hours the number of arrivals falls within 30.02 of the mean; in
approximately 99.7% of the hours the number of arrivals falls within 45.03 of the mean
4.37 a x = 47.71, s2 = 302.18 and s = 17.38

b.
250
200
Frequency
150
100
50
0
20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
AGE
c The histogram is approximately bell shaped allowing us to use the Empirical Rule.
Approximately 68% of adults are between 12.9 and 82.5 years old.
4.38a x = 77.86 and s = 85.35

b.
400
300
Frequency
200
100
0
0 50 100 150 200 250 300 More
TIME2
c. The histogram is positively skewed; we must use Chebysheff’s Theorem. At least 75% of
American adults watch between 0 and 249 minutes of television news.
110
4.39 a x = 23.4 and s = 19.6
b.
60
Frequency
40
20
0
10 20 30 40 50 60 70 80 90 100
AGECMEUS
The histogram is very positively skewed. As a result we can only use Chebysheff’s Theorem. At
least 75% of American born outside the United States were between 0 and 62.6 years old
25
4.40 First quartile: L25 = (15 + 1) = (16)(.25) = 4; the fourth number is 3.
100
50
Second quartile: L50 = (15 + 1) = (16)(.5) = 8; the eighth number is 5.
100
75
Third quartile: L75 = (15 + 1) = (16)(.75) = 12; the twelfth number is 7.
100
30
4.41 30th percentile: L30 = (10 + 1) = (11)(.30) = 3.3; the 30th percentile is 22.3.
100
80
80th percentile: L80 = (10 + 1) = (11)(.80) = 8.8; the 80th percentile 30.8.
100
20
4.42 20th percentile: L20 = (10 + 1) = (11)(.20) = 2.2; the 20th percentile is 43 + .2(51–43) = 44.6.
100
40
40th percentile: L40 = (10 + 1) = (11)(.40) = 4.4; the 40th percentile is 52 +.4(60–52) = 55.2.
100
25
4.43 First quartile: L25 = (13 + 1) = (14)(.25) = 3.5; the first quartile is 13.05.
100
50
Second quartile: L50 = (13 + 1) = (14)(.5) = 7; the second quartile is 14.7.
100
75
Third quartile: L75 = (13 + 1) = (14)(.75) = 10.5; the third quartile is 15.6.
100
111
30
4.44 Third decile: L 30 = (15 + 1) = (16)(.30) = 4.8; the third decile is 5 + .8(7 – 5) = 6.6.
100
60
Sixth decile: L60 = (15 + 1) = (16)(.60) = 9.6; the sixth decile is 17 + .6(18 – 17) = 17.6.
100
4.45 Interquartile range = 15.6 –13.05 = 2.55
4.46 Interquartile range = 7 – 3 = 4
4.47 First quartile = 5.75, third quartile = 15; interquartile range = 15 – 5.75 = 9.25
4.48
9 14 19 24 29 34
4.49 L85 = 75; The speed limit should be set at 75 mph.
4.50
0 50 100 150
a First quartile = 2, second quartile = 4, and third quartile = 8.

b Most executives spend little time reading resumes. Keep it short.
112
4.51 Dogs: First quartile = 1097.5, second quartile = 1204, and third quartile = 1337.
479 679 879 1079 1279 1479 1679 1879
Cats: First quartile = 743, second quartile = 856, and third quartile = 988.
479 679 879 1079 1279 1479 1679 1879
Dogs cost more money than cats. Both sets of expenses are positively skewed.
4.52 First quartile = 50, second quartile = 125, and third quartile = 260. The amounts are positively skewed.
4.53 BA First quartile = 25,730, second quartile = 27,765, and third quartile = 29836
18719 23719 28719 33719 38719 43719 48719
113
BSc First quartile = 29,927, second quartile = 33,397, and third quartile = 36,745
18719 23719 28719 33719 38719 43719 48719
BBA First quartile = 31,316, second quartile = 34,284, and third quartile = 39,551
18719 23719 28719 33719 38719 43719 48719
Other First quartile = 28,254, second quartile = 29,951, and third quartile = 32,905
18719 23719 28719 33719 38719 43719 48719
The starting salaries of BA and other are the lowest and least variable. Starting salaries for BBA and BSc
are higher.
114
4.54 a
140.1 160.1 180.1 200.1 220.1 240.1
b The quartiles are 145.11, 164.17, and 175.18

c There are no outliers.
d The data are positively skewed. One-quarter of the times are below 145.11 and one-quarter are
above 175.18.
4.55a Private course: The quartiles are 145.11, 164.17, and 175.18
213 263 313 363
Public course: The quartiles are 279, 296, and 307
213 263 313 363
115
b The amount of time taken to complete rounds on the public course are larger and more variable
than those played on private courses.
4.56 a The quartiles are 26, 28.5, and 32

b the times are positively skewed.
4.57 The quartiles are 8081.81, 9890.48, and 11,692.92. One-quarter of mortgage payments are
less than $607.19 and one quarter exceed $909.38.
4.58 TIME1
0 200 400 600 800 1000 1200
TIME2
0 200 400 600 800 1000 1200
Americans spend more time watching news on television than reading news on the Internet.
116
4.59
17 27 37 47 57 67 77 87 97
4.60 EDUC
0 5 10 15
SPEDUC
0 5 10 15
The two sets of numbers are quite similar.
4.61The quartiles are 34, 47, 60
117
18 28 38 48 58 68 78 88 98
Ages are symmetric.
4.62 The quartiles are 1, 2, 4
0 5 10 15 20 25
The number of hours of television watching is highly positively skewed.
4.63 There is a negative linear relationship. The strength is unknown.
s xy − 150
4.64 a. r = = = −.7813
sxsy (16)(12)
There is a moderately strong negative linear relationship.
b. R2 = r2 = (− .7813)2 = .6104
61.04% of the variation in y is explained by the variation in x.
118
4.65a. xi yi x i2 yi2 x i yi
20 14 400 196 280
40 16 1600 256 640
60 18 3600 324 1080
50 17 2500 289 850
50 18 2500 324 900
55 18 3025 324 990
60 18 3600 324 1080
70 20 4900 400 1400
Total 405 139 22,125 2,437 7,220

n n n n n
2 2
∑x
i =1
i = 405 ∑y
i =1
i = 139 ∑x
i =1
i = 22,125 ∑y
i =1
i = 2,437 ∑x y
i =1
i i = 7,220
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑yi =1
xi
i =1
i ⎥
⎥ 1 ⎡ (405)(139) ⎤
s xy = ∑
n − 1 ⎢⎢ i =1
x i yi −
n ⎥ = 8 − 1 ⎢7,220 −
⎣ 8 ⎥ = 26.16
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
⎢ n ⎜ ∑
⎟ ⎥
(405) 2 ⎤
⎢ x i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢22,125 −
1 ⎡
s 2x = ∑ ⎥ = 231.7
n − 1 ⎢ i =1 n ⎥ 8 − 1 ⎢⎣ 8 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s x = s 2x = 231.7 = 15.22
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(139) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢2,437 −
1 ⎡
s 2y = ∑ ⎥ = 3.13
n − 1 ⎢ i =1 n ⎥ 8 − 1 ⎣⎢ 8 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s y = s 2y = 3.13 = 1.77
s xy 26.16
r= = = .. 9711
sxsy (15.22)(1.77)
R2 = r2 = .97112 = .9430
The covariance is 26.16, the coefficient of correlation is .9711 and the coefficient of determination
is .9430.
94.30% of the variation in expenses is explained by the variation in total sales.
119
s xy 26.16
b. b1 = = = .113
s 2x 231.7
x=
∑x i
=
405
= 50.63
n 8
y=
∑y i
=
139
= 17.38
n 8
b 0 = y − b1x = 17.38 – (.113)(50.63) = 11.66
The least squares line is

ŷ = 11.66 + .113x
The estimated variable cost is .113 and the estimated fixed cost is 11.66.
4.66 xi yi x i2 yi2 x i yi
40 77 1,600 5,929 3,080
42 63 1,764 3,969 2,646
37 79 1,369 6,241 2,923
47 86 2,209 7,396 4,041
25 51 625 2,601 1,276
44 78 1,936 6,084 3,432
41 83 1,681 6,889 3,403
48 90 2,304 8,100 4,320
35 65 1,225 4,225 2,275
28 47 784 2,209 1,316
Total 387 719 15,497 53,643 28,712
n n n n n
∑ x i = 387 ∑ y i = 719 ∑ x i2 = 15,497 ∑ y i2 = 53,643 ∑x y i i = 28,712

i =1 i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑y
i =1
xi
i =1
i ⎥
⎥ 1 ⎡ (387)(719) ⎤
a s xy =
n − 1 ⎢⎢ i =1∑x i yi −
n ⎥ = 10 − 1 ⎢28,712 −
⎣ 10 ⎥ = 98.52
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
1 ⎢
n ⎜ ∑
⎟ ⎥
(387) 2 ⎤
⎢ x i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢15,497 −
⎡
s 2x = ∑ ⎥ = 57.79
n − 1 ⎢ i =1 n ⎥ 10 − 1 ⎢⎣ 10 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
120
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(719) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢53,643 −
1 ⎡
s 2y = ∑ ⎥ = 216.32
n − 1 ⎢ i =1 n ⎥ 10 − 1 ⎣⎢ 10 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 98.52
b r= = = .8811
sxsy (57.79)(216.32)
c R2 = r2 = .88112 = .7763
s xy 98.52
d b1 = = = 1.705
s 2x 57.79
x=
∑x i
=
387
= 38.7
n 10
y=
∑y i
=
719
= 71.9
n 10
b 0 = y − b1x = 71.9 – (1.705)(38.7) = 5.917
The least squares line is
ŷ = 5.917 + 1.705x
e. There is a strong positive linear relationship between marks and study time. For each additional
hour of study time marks increased on average by 1.705.
4.67 xi yi x i2 yi2 x i yi
599 9.6 358,801 92.16 5750.4
689 8.8 474,721 77.44 6063.2
584 7.4 341,056 54.76 4321.6
631 10.0 398,161 100.00 6310.0
594 7.8 352,836 60.84 4632.2
643 9.2 413,449 84.64 5915.6
656 9.6 430,336 92.16 6297.6
594 8.4 352,836 70.56 4989.6
710 11.2 504,100 125.44 7952.0
611 7.6 373,321 57.76 4643.6
593 8.8 351,649 77.44 5218.4
683 8.0 466,489 64.00 5464.0
Total 7,587 106.4 4,817,755 957.2 67,559.2
n n n n n
∑ x i =7,587 ∑ y i = 106.4 ∑ x i2 = 4,817,755 ∑ y i2 = 957.2 ∑x y i i =

i =1 i =1 i =1 i =1 i =1
67,559.2
121
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑
i =1
xi
i =1
y i ⎥
⎥ 1 ⎡ (7,587)(106.4) ⎤
s xy = ∑
n − 1 ⎢⎢ i =1
x i yi −
n ⎥ = 12 − 1 ⎢67,559.2 −
⎣ 12 ⎥ = 26.16
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
1 ⎢ n ⎜
⎝ i 1
∑ ⎟ ⎥
⎠
1 ⎡
⎢ 4,817, 755 −
(7,587) 2 ⎤
⎥ = 1,897.7
s 2x = ⎢ x i2 − =
∑
n − 1 ⎢ i =1 n
⎥ = 12 − 1 ⎢
⎥ ⎣ 12 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s x = s 2x = 1,897.7 = 43.56
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎟ ⎥
1 ⎢ n ⎜
⎝ i 1
∑ ⎟ ⎥
⎠
1 ⎡
⎢957.2 −
(106.4) 2 ⎤
⎥ = 1.25
s 2y = ⎢ y i2 − =
∑
n − 1 ⎢ i =1 n
⎥ = 12 − 1 ⎢
⎥ ⎣ 12 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s Y = s 2Y = 1.25 = 1.12
s xy 26.16
r= = = .5362
sxsy (43.56)(1.12)
R2 = r2 = .53622 = .2875
The covariance is 26.16, the coefficient of correlation is .5362, and the coefficient of
determination is .2875. The coefficient of determination tells us that 28.75% of the variation in
MBA GPAs is explained by the variation in GMAT scores.
4.68
R2 = r2 = (−.6332)2 = .4009; 40.09% of the variation in the employment rate is explained by the
variation in the unemployment rate.
4.69 a
122
R2 = r2 = (.2543)2 = .0647.
b There is a weak linear relationship between age and medical expenses. Only 6.47% of the
variation in average medical bills is explained by the variation in age.
c
80.00
70.00
60.00 y = 0 .2257x -‐ 5.9662
R² = 0 .0647
50.00
Expense
40.00
30.00
20.00
10.00
0.00
-‐10.00 0 20 40 60 80 100
Age
The least squares line is ŷ = −5.966 + .2257x
d For each additional year of age mean medical expenses increase on average by $.2257 or 23
cents.
e Charge 25 cents per day per year of age.
4.70
R2= (−.2435)2 = .0593
Only 5.93% of the variation in the number of houses sold is explained by the variation in interest
rates.
123
4.71
300
250
200
Wells drilled
y = 0.1938x + 71.344
150 R² = 0.0055
100
50
0
0 20 40 60 80 100 120 140 160
Price of oil
Only 0.55% of the variation in the number of wells drilled is explained by the variation in the
price of oil. The relationship is too weak to interpret the value of the slope coefficient.
4.72
R2 = (.0830)2 = .0069.
There is a very weak positive relationship between the two variables.
124
4.73
480
460
440
420
Labor cost
400
y = 3 .3x + 3 15.5
380 R² = 0 .5925
360
340
320
300
0 10 20 30 40 50
Batch size
ŷ = 315.5 + 3.3x; Fixed costs = $315.50, variable costs = $3.30
4.74
1800
1600
1400 y = 71.654x + 263.4
R² = 0.5437
1200
1000
Cost
800
600
400
200
0
0 2 4 6 8 10 12 14 16
Time
ŷ = 263.4 + 71.65x; Estimated fixed costs = $263.40, estimated variable costs = $71.65
125
4.75a
50,000
45,000
y = 510.37x -‐ 11028
40,000 R² = 0.4641
35,000
Home attendnace
30,000
25,000
20,000
15,000
10,000
5,000
0
0 20 40 60 80 100 120
Wins
b The slope coefficient is 510.37; home attendance increases on average by 510.37 for each win.
46.41% of the variation in home attendance is explained by the variation in the number of wins.
4.76a
40,000
35,000 y = 58.59x + 25588
R² = 0.0915
30,000
Away attendance
25,000
20,000
15,000
10,000
5,000
0
0 20 40 60 80 100 120
Wins
R2 = .0915; there is a very weak relationship between the two variables.

b The slope coefficient is 58.59; away attendance increases on average by 58.59 for each win.
However, the relationship is very weak.
126
4.77
70
60
y = 0.26x + 22.287
50
R² = 0.0411
40
Payroll
30
20
10
0
0.000 20.000 40.000 60.000 80.000 100.000 120.000
Wins
a. The slope coefficient is .26; for each million dollars in payroll the number of wins increases on
average by .26. Thus, to cost of winning one addition game is 1/.26 million = $3.846 million.
b. The coefficient of determination tells us that only 4.11.9% of the variation in the number of
wins is explained by the variation in payroll,
4.78
16
14 y = 0.0428x + 3.3651
R² = 0.0866
12
10
Payroll
8
6
4
2
0
0 50 100 150 200 250
Wins
127
a. The slope coefficient is .0428; for each million dollars in payroll the number of wins increases
on average by .0428. Thus, to cost of winning one addition game is 1/.0428 million = $23.364
million.
b. The coefficient of determination = .0866, which reveals that the linear relationship is very weak.
4.79
60
y = 0.1526x + 28.559
50 R² = 0.0876
40
Payroll
30
20
10
0
0.000 20.000 40.000 60.000 80.000 100.000
Wins
a. The slope coefficient is .1526; for each million dollars in payroll the number of wins increases
on average by .1526. Thus, to cost of winning one addition game is 1/.1526 million = $6.553
million.
b. The coefficient of determination = .0876, which reveals that the linear relationship is very weak.
128
4.80a
25,000
Home attendance 20,000
15,000
y = 8 4.391x + 1 4037
10,000 R² = 0 .2468
5,000
0
0 10 20 30 40 50 60 70
Wins
For each additional win home attendance increases on average by 84.391. The coefficient of
determination is .2468; there is a weak relationship between the number of wins and home
attendance.
19,500
19,000 y = 3 1.151x + 1 6228

R² = 0 .4407
18,500
Away attendance
18,000
17,500
17,000
16,500
16,000
0 10 20 30 40 50 60 70
Wins
For each additional win away attendance increases on average by 31.151. The coefficient of
determination is .4407; there is a moderately strong relationship between the number of wins and
away attendance.
129
4.81
120
Percent of capacity 100
80
y = 0 .4618x + 7 1.98
60 R² = 0 .4023
40
20
0
0 10 20 30 40 50 60 70
Wins
R2 = .4023. The relationship between wins and home attendance as a percentage of capacity is
weaker than the relationship between wins and home attendance.
4.82
100,000
90,000
80,000
70,000
Home attendance
60,000
50,000
40,000 y = 947.38x + 59919
R² = 0.1108
30,000
20,000
10,000
0
0 2 4 6 8 10 12 14 16
Wins
For each additional win home attendance increases on average by 947.38. The coefficient of
determination is .1108; there is a very weak linear relationship between the number of wins and
home attendance.
130
80,000
70,000
60,000
Away attendance
50,000 y = 216.74x + 65775
40,000 R² = 0.0322
30,000
20,000
10,000
0
0 2 4 6 8 10 12 14 16
Wins
For each additional win away attendance increases on average by 216.74. The coefficient of
determination is .0322; there is a very weak linear relationship between the number of wins and
away attendance.
4.83
Pct of Capacity

120
100
Percent of capacity
80
y = 1.5469x + 83.015
60 R² = 0.3304
40
20
0
0 2 4 6 8 10 12 14 16
Wins
R2 = .3304. The relationship between wins and home attendance as a percentage of capacity is
stronger than the relationship between wins and home attendance.
131
4.84 a
30
25
20
y = -‐0.2098x + 5.7928
TVHOURS
15 R² = 0.0572
10
0
0 5 10 15 20 25
EDUC
There is a weak negative linear relationship between education and television watching.
b R2 = .0572; 5.72% of the variation in the amount of television is explained by the variation in
education.
4.85 Correlation matrix
4.86 Correlation matrix
132
4.87 b1 R2
AT&T 0.687 .318
Aetna 1.256 .296
Cigna 1.829 .463
Coca-Cola 0.601 .324
Disney 1.104 .592
Ford 2.654 .296
McDonald’s 0.637 .314
4.88 b1 R2
Barrick Gold 0.594 .071
Bell Canada Enterprises (BCE) 0.399 .089
Bank of Montreal (BMO) 0.610 .164
Enbridge 0.314 .109
Fortis 0.211 .032
Methanex 1.301 .270
Research in Motion (RIM) 1.465 .201
Telus 0.446 .097
Trans Canada Pipeline 0.393 .197
4.89 b1 R2
Amazon 1.324 .267
Amgen 0.492 .096
Apple 1.358 .401
Cisco Systems 1.100 .604
Google 1.075 .327
Intel 1.074 .556
Microsoft 0.865 .436
Oracle 0.866 .526
Research in Motion 1.920 .387
133
4.90 a
b We can see that among those who repaid the mean score is larger than that of those who did not
and the standard deviation is smaller. This information is similar but more precise than that
obtained in Exercise 3.23.
4.91 Repaid loan:
419 519 619 719 819 919
134
Defaulted on loan:
419 519 619 719 819 919
The box plots make it a little easier to see the overlap between the two sets of data (indicating that
the scorecard is not very good).
4.92
R2 = .67842 = .4603; 46.03% of the variation in statistics marks is explained by the variation in
calculus marks. The coefficient of determination provides a more precise indication of the
strength of the linear relationship.
4.93
2500
y = 116.53x -‐ 369.93

2000 R² = 0.9334
1500
Cost
1000
500
0
0 5 10 15 20 25
Speed
135
The least squares line is ŷ = 369.93 + 116.53x. On average for each addition mph the cost of
repair increases by $116.53.
4.94
90 y = 0 .6041x + 1 7.933
R² = 0 .0505
80
70
60
50
Income
40
30
20
10
0
60 65 70 75 80
Height
a ŷ = 17.933 + .6041x
b The coefficient of determination is .0505, which indicates that only 5.05% of the variation in
incomes is explained by the variation in heights.
4.95
3000.00
y = 19.059x + 1087.7
2500.00 R² = 0.0779
2000.00
Sales
1500.00
1000.00
500.00
0.00
0 5 10 15 20 25 30 35
Time
136
The coefficient of determination is .0779, which indicates that only 7.79% of the variation in sales
is explained by the time between movies.
4.96a
400
350 y = 0 .07x + 1 03.44
R² = 0 .5201
300
250
Price
200
150
100
50
0
0 500 1000 1500 2000 2500 3000 3500 4000
Size
b. The slope coefficient is .07; For each additional square foot the price increases on average by
$.07 thousand. More simply for each additional square foot the price increases on average by$70.
c. From the least squares line we can more precisely measure the relationship between the two
variables.
4.97 B.A.
80
60
Frequency
40
20
0
25000 30000 35000 40000 45000 50000
BA
137
B.Sc.
20
15
Frequency
10
5
0
25000 30000 35000 40000 45000 50000
Salary offer
B.B.A.
40
30
Frequency
20
10
0
25000 30000 35000 40000 45000 50000
Salary offer
Other
40
30
Frequency
20
10
0
25000 30000 35000 40000 45000 50000
Salary offer
Using the same class limits the histograms provide more detail than do the box plots.
138
4.98 Private course
50
40
Frequency 30
20
10
0
220 230 240 250 260
Time
Public course
50
40
Frequency
30
20
10
0
250 275 300 325 350 375
Time
The information obtained here is more detailed than the information provided by the box plots.
4.99
a x = 35.01, median = 36
139
b s = 7.68
c Half of the bone density losses lie below 36. At least 75% of the numbers lie between 19.64 and
50.38, at least 88.9% of the numbers lie between 11.96 and 58.06.
4.100
a x = 29,913, median = 30,660
b s 2 = 148,213,791; s = 12,174
-‐10 9990 19990 29990 39990 49990 59990
d The number of coffees sold varies considerably.
140
4.101
R2 = r2 = .57422 = .3297; 32.97% of the variation in bone loss is explained by the variation in age.
4.102 a & b
70000
60000 y = -‐553.7x + 49337

R² = 0.5489
50000
40000
Coffees
30000
20000
10000
0
-‐20 0 20 40 60 80
Temperatyre
R2 = .5489 and the least squares line is ŷ = 49,337 – 553.7x
c 54.8% of the variation in the number of coffees sold is explained by the variation in temperature.
For each additional degree of temperature the number of coffees sold decreases on average by 554
cups. Alternatively for each 1-degree drop in temperature the number of coffees increases on
average, by 553.7 cups.
d We can measure the strength of the linear relationship accurately and the slope coefficient gives
information about how temperature and the number of coffees sold are related.
4.103a mean, median, and standard deviation
141
b
x = 93.90, s = 7.72
c We hope Chris is better at statistics than he is golf.
4.104
a x = 26.32 and median = 26
b s 2 = 88.57, s = 9.41
142
c.
2 12 22 32 42 52
d The times are positively skewed. Half the times are above 26 hours.
4.105
120
y = 1.5176x + 39.602
110 R² = 0.8021
100
Total score
90
80
70
60
20 25 30 35 40 45 50 55
Putts
80.21% of the variation in scores is explained by the variation in the number of putts.
143
4.106 a & b
60
50 y = 3.146x -‐ 8.2897

R² = 0.412
40
Internet use
30
20
10
0
0 5 10 15 20
Education
R2 = .412 and the least squares line is ŷ = −8.2897 + 3.146x
c 41.2% of the variation in Internet use is explained by the variation in education. For each
additional year of education Internet use increases on average by 3.146 hours.
information about how education and Internet use are related.
4.107
x = 150.77, median = 150.50, and s = 19.76. The average crop yield is 150.77 and there is a great
deal of variation from one plot to another.
144
4.108a & b
250
200 y = 0 .128x + 8 9.543

R² = 0 .3692
150
Corn
100
50
0
0 100 200 300 400 500 600 700 800
Rainfall
R2 = .369 and the least squares line is ŷ = 89.543 + .128 Rainfall
c 36.92% of the variation in yield is explained by the variation in rainfall. For each additional
inch of rainfall yield increases on average by .128 bushels.
information about how rainfall and crop yield are related.
4.109
250
y = 0.1802x + 120.37
200 R² = 0.1549
150
Corn
100
50
0
0 50 100 150 200 250 300
Fertilizer
R2 = .1549 and the least squares line ŷ = 120.37 + .1802 Fertilizer
145
c 15.49% of the variation in yield is explained by the variation in the amount of fertilizer. For
each additional unit of fertilizer yield increases on average by.180 bushels.
information about how the amount of fertilizer and crop yield are related.
4.110a
b The mean debt is $12,067. Half the sample incurred debts below $12,047 and half incurred debts
above. The mode is $11,621.
Case 4.1 a Scatter diagrams with time as the independent variable and temperature anomalies as
the dependent variable
2.5
y = 0 .0006x -‐ 0.4754
2
R² = 0 .3708
1.5
1
0.5
0
-‐0.5 0 200 400 600 800 1,000 1,200 1,400 1,600
-‐1
-‐1.5
-‐2
-‐2.5
Month
146
Monthly average increase is .0006. For the 1600 month period the increase was 1600(.0006) = .96o
Celsius.
Scatter diagrams with carbon dioxide levels as the independent variable and temperature
anomalies as the dependent variable
1.5 y = 0 .0151x -‐ 4.9604

R² = 0 .5075
0.5
-‐0.5
-‐1
-‐1.5
300 320 340 360 380 400
CO2
The coefficient of determination is .5075, which means that 50.75% of the variation in
temperature anomalies is explained by the variation in CO2levels. There is a moderately strong
linear relationship.
Case 4.21880 to 1940
1.5
y = 0 .0007x -‐ 0.4776
1 R² = 0 .1919
0.5
0
-‐0.5
-‐1
-‐1.5
-‐2
-‐2.5
0 100 200 300 400 500 600 700 800
Month
From 1880 to 1940 the earth warned at an average monthly rate of .0007o Celsius.
147
1941 to 1975
Temperature anomalies 0.5
-‐0.5
-‐1
y = -‐0.0004x + 0.0332
R² = 0.0251
-‐1.5
0 100 200 300 400 500
Month
From 1941 to 1975 the earth cooled at an average monthly rate of .0004o Celsius
1976 to 1997
2
y = 0 .0021x -‐ 0.0581
1.5 R² = 0 .197
0.5
-‐0.5
-‐1
-‐1.5
0 50 100 150 200 250 300
Month
From 1976 to 1997 the earth warmed at an average monthly rate of .0021o Celsius.
148
1998 to 2009
2 y = 0.0012x + 0.6932
R² = 0.0241
Temperature anomalies 1.5
0.5
-‐0.5
0 20 40 60 80 100 120 140 160
Month
From 1998 to 2009 the earth warmed at an average monthly rate of .0012o Celsius
Over different periods of time the earth has warmed and cooled.
Case 4.3 2003-04 Season
60
y = 0.1526x + 28.559
50 R² = 0.0876
40
Wins
30
20
10
0
$0 $20 $40 $60 $80 $100
Payroll ($millions)
The cost of winning one additional game is 1million/.1526 = $6.553 million. However, the
coefficient of determination is only .0876, which tells us that there are many other variables that
determine how well a team will do.
149
2005-06 Season
70
y = 0 .7795x + 1 4.256
60
R² = 0 .3072
50
40
Wins
30
20
10
0
$0 $10 $20 $30 $40 $50
Payroll ($millions)
The cost of winning one additional game is 1million/.7795 = $1.283 million. The coefficient of
determination is .3072.
The small coefficient of determination in the year before the strike seems to indicate that team
owners were spending large amounts of money and getting little in return. The results are
markedly different in the year after the strike. There is a much stronger linear relationship between
payroll and the number of wins and the cost of winning one additional game is considerably
smaller.
Case 4.4
The coefficient of determination is (−.1787)2 = .0319. There is a weak negative linear relationship
between percentage of rejected ballots and Percentage of “yes” votes.
150
The coefficient of determination is (.3600)2 = .1296. There is a moderate positive linear
relationship between percentage of rejected ballots and Percentage of Allophones.
The coefficient of determination is (.0678)2 = .0046. There is a very weak positive linear
relationship between percentage of rejected ballots and Percentage of Allophones.
The statistics provide some evidence that electoral fraud has taken place.
151
152
Chapter 5
5.1 In an observational study, there is no attempt to control factors that might influence the
variable of interest. In an experimental study, a factor (such as regular use of a fitness center) is
controlled by randomly selecting who is exposed to that factor, thereby reducing the influence of
other factors on the variable of interest.
5.2a The study is observational. The statistics practitioner did not randomly assign stores to buy
cans or bottles.
b Randomly assign some stores to receive only cans and others to receive only bottles.
5.3 Randomly sample smokers and nonsmokers and compute the proportion of each group that has
lung cancer.
b The study is observational. Experimental data would require the statistics practitioner to
randomly assign some people to smoke and others not to smoke.
5.4a A survey can be conducted by means of a personal interview, a telephone interview, or a self-
administered questionnaire.
b A personal interview has a high response rate relative to other survey methods, but is expensive
because of the need to hire well-trained interviewers and possibly pay travel-related costs if the
survey is conducted over a large geographical area. A personal interview also will likely result in
fewer incorrect responses that arise when respondents misunderstand some questions. A telephone
interview is less expensive, but will likely result in a lower response rate. A self-administered
questionnaire is least expensive, but suffers from lower response rates and accuracy than
interviews.
5.5 Five important points to consider when designing a questionnaire are as follows:
(1) The questionnaire should be short.
(2) Questions should be short, clearly worded, and unambiguous.
(3) Consider using dichotomous or multiple-choice questions, but take care that respondents
needn’t make unspecified assumptions before answering the questions.
(4) Avoid using leading questions.
(5) When preparing the questions, think about how you intend to tabulate and analyze the
responses.
5.6a The sampled population will exclude those who avoid large department stores in favor or
smaller shops, as well as those who consider their time too valuable to spend participating in a
153
survey. The sampled population will therefore differ from the target population of all customers
who regularly shop at the mall.
b The sampled population will contain a disproportionate number of thick books, because of the
manner in which the sample is selected.
c The sampled population consists of those eligible voters who are at home in the afternoon,
thereby excluding most of those with full-time jobs (or at school).
5.7a The Literary Digest was a popular magazine in the 1920s and 1930s which had correctly
predicted the outcome of many presidential elections. To help predict the outcome of the 1936
presidential election, the Literary Digest mailed sample ballots to 10 million prospective voters.
Based on the results of the ballots returned, the magazine predicted that the Republican candidate,
Alfred Landon, would defeat the Democratic incumbent, Franklin D. Roosevelt, by a 3 to 2
margin. In fact, Roosevelt won a landslide victory, capturing 62% of the votes.
b The main reason for the poll being so wrong was nonresponse bias resulting from a self-selected
sample, causing the sample to be unrepresentative of the target population. (Only 2.3 million
ballots were returned.) The second reason was selection bias, resulting from poor sampling design,
causing the sampled population and the target population to differ. Most of those to whom a ballot
was sent were selected from the Literary Digest’s subscription list and from telephone directories.
These people tended to be wealthier than average and tended to vote Republican.
5.8a A self-selected sample is a sample formed primarily on the basis of voluntary inclusion, with
little control by the designer of the survey.
b Choose any recent radio or television poll based on responses of listeners who phone in on a
volunteer basis.
c Self-selected samples are usually biased, because those who participate are more interested in
the issue than those who don’t, and therefore probably have a different opinion.
5.9 We should ignore the results because this is an example of a self-selected sample.
5.10 No, because the sampled population consists of the responses about the professor’s
course. We cannot make draw inferences about all courses.
5.11 We used Excel to generate 40 three-digit random numbers. Because we will ignore all
randomly generated numbers over 800, we can expect to ignore about 20% (or about 8 to 10) of
the randomly generated numbers. We will also ignore any duplication. We therefore chose to
generate 40 three-digit random numbers, and will use the first 25 unique random numbers less
than 801 to select our sample. The 40 numbers generated are shown below, with a stroke through
154
those to be ignored.
6 357 456 449 862 154 55 412 475 430

999 912 60 207 717 651 10 294 327 165
576 871 990 354 390 540 893 181 496 870
738 820 32 963 160 32 231 86 970 46
5.12 We used Excel to generate 30 six-digit random numbers. Because we will ignore any
duplicate numbers generated, we generated 30 six-digit random numbers and will use the first 20
unique
random numbers to select our sample. The 30 numbers generated are shown below.
169,470 744,530 22,554 918,730 320,262 503,129

318,858 698,203 822,383 938,262 800,806 56,643
836,116 123,936 80,539 154,211 391,278 940,154
110,630 856,380 222,145 692,313 949,828 561,511
909,269 811,274 288,553 749,627 858,944 39,308
5.13 Stratified random sampling is recommended. The strata are the school of business, the faculty
of arts, the graduate school and the all the other schools and faculties would be the fourth stratum.
The data can be used to acquire information about the entire campus but also compare the four
strata.
5.14 A stratified random sampling plan accomplishes the president’s goals. The strata are the four
areas enabling the statistics practitioner to learn about the entire population but also compare the
four areas.
5.15 The operations manager can select stratified random samples where the strata are the four
departments. Simple random sampling can be conducted in each department.
5.16 Use cluster sampling, letting each city block represent a cluster.
5.17a Sampling error refers to an inaccuracy in a statement about a population that arises because
the statement is based only on sample data. We expect this type of error to occur because we are
making a statement based on incomplete information. Nonsampling error refers to mistakes made
in the acquisition of data or due to the sample observations being selected improperly.
b Nonsampling error is more serious because, unlike sampling error, it cannot be diminished by
taking a larger sample.
155
5.18 Three types of nonsampling errors:
(1) Error due to incorrect responses
(2)Nonresponse error, which refers to error introduced when responses are not obtained from some
members of the sample. This may result in the sample being unrepresentative of the target
population.
(3)Error due to selection bias, which arises when the sampling plan is such that some members of
the target population cannot possibly be selected for inclusion in the sample.
5.19 Yes. A census will likely contain significantly more nonsampling errors than a carefully
conducted sample survey.
156
Chapter 6
6.1 a Relative frequency approach
b If the conditions today repeat themselves an infinite number of days rain will fall on 10% of the
next days.
6.2 a Subjective approach

b If all the teams in major league baseball have exactly the same players the New York Yankees
will win 25% of all World Series.
6.3 a {a is correct, b is correct, c is correct, d is correct, e is correct}

b P(a is correct) = P(b is correct) = P(c is correct) = P(d is correct) = P(e is correct) = .2
c Classical approach
d In the long run all answers are equally likely to be correct.
6.4 a Subjective approach

b The Dow Jones Industrial Index will increase on 60% of the days if economic conditions remain
unchanged.
6.5 a P(even number) = P(2) + P(4) + P(6) = 1/6 + 1/6 + 1/6 = 3/6 = 1/2
b P(number less than or equal to 4) = P(1) + P(2) + P(3) + P(4) = 1/6 + 1/6 + 1/6 +1/6 = 4/6 = 2/3
c P(number greater than or equal to 5) = P(5) + P(6) = 1/6 + 1/6 = 2/6 = 1/3
6.6 {Adams wins. Brown wins, Collins wins, Dalton wins}
6.7a P(Adams loses) = P(Brown wins) + P(Collins wins) + P(Dalton wins) = .09 + .27 + .22 = .58
b P(either Brown or Dalton wins) = P(Brown wins) + P(Dalton wins) = .09 + .22 = .31
c P(either Adams, Brown, or Collins wins) = P(Adams wins) + P(Brown wins) + P(Collins wins)
= .42 + .09 + .27 = .78
6.8 a {0, 1, 2, 3, 4, 5}
b {4, 5}
c P(5) = .10
d P(2, 3, or 4) = P(2) + P(3) + P(4) = .26 + .21 + .18 = .65
e P(6) = 0
6.9 {Contractor 1 wins, Contractor 2 wins, Contractor 3 wins}
157
6.10 P(Contractor 1 wins) = 2/6, P(Contractor 2 wins) = 3/6, P(Contractor 3 wins) = 1/6
6.11 a {Shopper pays cash, shopper pays by credit card, shopper pays by debit card}
b P(Shopper pays cash) = .30, P(Shopper pays by credit card) = .60, P(Shopper pays by debit card)
= .10
c Relative frequency approach
6.12 a P(shopper does not use credit card) = P(shopper pays cash) + P(shopper pays by debit card)
= .30 + .10 = .40
b P(shopper pays cash or uses a credit card) = P(shopper pays cash) + P(shopper pays by credit
card)
= .30 + .60 = .90
6.13 {single, divorced, widowed}
6.14 a P(single) = .15, P(married) = .50, P(divorced) = .25, P(widowed) = .10

b Relative frequency approach
6.15 a P(single) = .15

b P(adult is not divorced) = P(single) + P(married) + P(widowed) = .15+ .50 + .10 = .75
c P(adult is either widowed or divorced) = P(divorced) + P(widowed) = .25 + .10 = .35
6.16 P( A1 ) = .1 + .2 = .3, P( A 2 ) = .3 + .1 = .4, P( A 3 ) = .2 + .1 = .3.
P( B1 ) = .1 + .3 + .2 = .6, P( B 2 ) = .2 + .1 + .1 = .4.
6.17 P( A1 ) = .4 + .2 = .6, P( A 2 ) = .3 + .1 = .4. P( B1 ) = .4 + .3 = .7, P( B 2 ) = .2 + .1 = .3.
P(A 1 and B1 ) .4
6.18 a P(A 1 | B1 ) = = = .57
P ( B1 ) .7
P(A 2 and B1 ) .3
b P(A 2 | B1 ) = = = .43
P(B1 ) .7
c Yes. It is not a coincidence. Given B1 the events A1 and A 2 constitute the entire sample space.
P(A 1 and B 2 ) .2
6.19 a P(A 1 | B 2 ) = = = .67
P(B 2 ) .3
158
P(A1 and B2 ) .2
b P( B 2 | A 1 ) = = = .33
P(A1) .6
c One of the conditional probabilities would be greater than 1, which is not possible.
6.20 The events are not independent because P(A1 | B 2 ) ≠ P(A1 ) .
6.21 a P( A 1 or B1 ) = P(A1 ) + P(B1 ) − P(A1 and B1 ) = .6 + .7 - .4 = .9
b P( A 1 or B 2 ) = P(A1 ) + P(B 2 ) − P(A1 and B 2 ) = .6 + .3 - .2 = .7
c P( A 1 or A 2 ) = P(A1 ) + P(A 2 ) = .6 + .4 = 1
P(A1 and B1 ) .20

6.22 P(A1 | B1 ) = = = .25 ; P(A1 ) = .20 + .05 = .25 ; the events are
P(B1 ) .20 + .60
independent.
P(A1 and B1 ) .20

6.23 P(A1 | B1 ) = = = .571; P(A1 ) = .20 + .60 = .80; the events are
P(B1 ) .20 + .15
dependent.
6.24 P( A1 ) = .15 + .25 = .40, P( A 2 ) = .20 + .25 = .45, P( A 3 ) = .10 + .05 = .15.
P( B1 ) = .15 + 20 + .10 = .45, P( B 2 ) = .25 + .25 + .05 = .55.
P(A 2 and B 2 ) .25

6.25 a P( A 2 | B 2 ) = = = .455
P( B 2 ) .55
P(A 2 and B 2 ) .25
b P( B 2 | A 2 ) = = = .556
P( A 2 ) .45
P(A 2 and B1 ) .20
c P( B1 | A 2 ) = = = .444
P(A 2 ) .45
6.26 a P( A 1 or A 2 ) = P( A 1 ) + P( A 2 ) = .40 + .45 = .85
b P( A 2 or B 2 ) = P( A 2 ) + P( B 2 ) – P( A 2 and B 2 ) = .45 + .55 - .25 = .75
c P( A 3 or B1 ) =P( A 3 ) + P( B1 ) – P( A 3 and B1 ) = .15 + .45 - .10 = .50
P(promoted and female) .03

6.27 a P(promoted | female) = = = .20
P(female) .03 + .12
P(promoted and male) .17
b P(promoted | male) = = = .20
P(male) .17 + .68
159
c No, because promotion and gender are independent events.
6.28 a P(debit card) = .04 + .18 + .14 = .36

P(credit card and over $100 .23
b P(over $100 | credit card) = = = .49
P(credit card) .03 + .21 + .23
c P(credit card or debit card) = P(credit card) + P(debit card) = .47 + .36 = .83
6.29 a P(Less than high school) = .077 + .110 = .187

b P(college/university | female) =
P(college / university and female) .092
= = .198
P(female) .077 + .154 + .141 + .092
P(high school and male) .201

c b P(high school | male) = = = .566
P(male) .154 + .201
6.30 a P(He is a smoker) = .12 + .19 = .31

b P(He does not have lung disease) = .19 + .66 = .85
c P(He has lung disease | he is a smoker) =
P(he has lung disease and he is a smo ker) .12
= = .387
P(he is a smo ker) .31
d P(He has lung disease | he does not smoke) =
P(he has lung disease and he does not smoke) .03
= = .043
P(he does not smoke) .69
6.31 The events are dependent because P(he has lung disease) = .15, P(he has lung disease | he is a
smoker) = .387
P(manual and math − stats) .23

6.32 a P(manual | math-stats) = = = .390
P(math − stats) .23 + .36
b P(computer) = .36 + .30 = .66
c No, because P(manual) = .23 + .11 = .34, which is not equal to P(manual | math-stats).
6.33 a P(customer will return and good rating) =.35

P(good rating and will return ) .35 .35
b P(good rating | will return) = = = = .538
P(will return ) .02 + .08 + .35 + .20 .65
P(good rating and will return ) .35 .35

c P(will return| good rating) = = = .714
P(good rating) .35 + .14 .49
d (a) is the joint probability and (b) and (c) are conditional probabilities
160
6.34 a P(ulcer) = .01 + .03 + .03 + .04 = .11
P(ulcer and none) .01 .01
b P(ulcer | none) = = = = .043
P(none) .01 + .22 .23
P(ulcer and none) .01 .01

c P(none | ulcer) = = = = .091
P(ulcer) .01 + .03 + .03 + .04 .11
P(ulcer and none)

d P(One, two, or more than two | no ulcer) = 1 − = 1 − .091 = .909
P(ulcer)
P(Insufficient work and 25 − 54) .180

6.35 a P(Insufficient work | 25-54) = = = .252
P(25 − 54) .320 + .180 + .214
b P(65 and over) = .029 + .011 + .016 = .056
c P(65 and over |plant or company closed or moved) =
P(65 and over and plant or company closed or moved) .029
= = .064
P(plant or company closed or moved) .015 + .320 + ..089 + .029
6.36 a P(remember) = .15 + .18 = .33

P(remember and violent) .15 .15
b P(remember | violent) = = = = .30
P(violent) .15 + .35 .50
c Yes, the events are dependent.
P(above average and murderer) .27 .27

6.37 a P(above average | murderer) = = = = .563
P(murderer) .27 + .21 .48
b No, because P(above average) = .27 + .24 = .51, which is not equal to P(above average
testosterone | murderer).
6.38a P(Health insurance) = .167+.209+.225+.177 = .778

b. P(Person 55-64 | No health insurance)
P(Person 55 − 64 and No health insurance) .026 .026
= = = = .128
P(Person 55 − 64) .177 + .026 .203
c. P(Person 25-34|No health insurance) =
P(Person 25 − 34 and No health insurance) .085 .085
= = = .385
P( No health insurance) .085 + .061 + .049 + .026 .221
P(Violent crime and primary school) .393 .393

6.39a = = = .673
P(Pr imary school) .393 + .191 .584
b. P(No violent crime) = .191 + .010 + .007 + .015 = .223
161
P(Violent crime and enrollment less than 300) .159 .159
6.40a = = = .636
P(Enrollment less than 300) .159 + .091 .250
P(Violent crime and enrollment less than 300) .159 .159

b. = = = .205
P(Violent crime) .159 + .221 + .289 + .108 .777
P(new and overdue) .06 .06

6.41 a P(new | overdue) = = = = .103
P(overdue) .06 + .52 .58
P(new and overdue) .06 .06

b P(overdue | new) = = = = .316
P(new ) .06 + .13 .19
c Yes, because P(new) = .19 ≠ P(new | overdue)
6.42 a P(under 20) = .464 + .147 + .237 = .848

b P(retail) = .237 + .035 + .005 = .277
P(20 to 99 and construction) .039 .039
c P(20 to 99 | construction) = = = = .077
P(construction) .464 + .039 + .005 .508
6.43 a P(fully repaid) = .19 + .64 = .83

P(fully repaid and under 400) .19 .19
b P(fully repaid | under 400) = = = = .594
P(under 400) .19 + .13 .32
P(fully repaid and 400 or more) .64 .64

c P(fully repaid | 400 or more) = = = = .941
P(400 or more) .64 + .04 .68
d No, because P(fully repaid) ≠ P(fully repaid | under 400)
P(purchase and see ad) .18 .18

6.44 P(purchase | see ad) = = = = .30; P(purchase) = .18 +
P(see ad) .18 + .42 .60
.12 = .30. The events are independent and thus, the ads are not effective.
6.45 a P(unemployed | high school graduate) =

P(unemployed and high school graduate) .014 .014
= = = .047
P(high school graduate) .282 + .014 .296
b P(employed) = .091 + .282 + .166 + .095 + .213 + .115 = .962
c P(advanced degree | unemployed) =
P(advanced deg ree and unemployed) .002 .002
= = = .053
P(unemployed) .008 + .014 + .007 + .003 + .004 + .002 .038
d P(not a high school graduate) = .091 + .008 = .099
6.46 a P(bachelor’s degree | west)
162
P(bachelor' s deg ree and west) .046 .046
= = = = .201
P(west) .036 + .059 + .045 + .020 + .046 + .023 .229
b P(northeast | high school graduate)
P(northeast and high school graduate) .063 .063
= = = = .199
P(high school graduate) .063 + .078 + .117 + .059 .317
c P(south) = .059 + .117 + .061 + .030 + .065 + .032 = .364
d P(not south) = 1 –P(south) = 1−.364 = .636
6.47
6.48
163
6.49
6.50
6.51
164
6.52
a P(R and R) = .81

b P(L and L) = .01
c P(R and L) + P(L and R) = .09 + .09 = .18
d P(Rand L) + P(L and R) + P(R and R) = .09 + .09 + .81 = .99
6.53 a & b
c 0 right-handers 1
1 right-hander 3
2 right-handers 3
3 right-handers 1
d P(0 right-handers) = .001
165
P(1 right-hander) = 3(.009) = .027
P(2 right-handers) = 3(.081) = .243
P(3 right-handers) = .729
6.54a
b P(RR) = .8091
c P(LL) = .0091
d P(RL) + P(LR) = .0909 + .0909 = .1818
e P(RL) + P(LR) + P(RR) = .0909 + .0909 + .8091 = .9909
6.55a
P(0 right-handers) = (10/100)(9/99)(8/98) = .0007
166
P(1 right-hander) = 3(90/100)(10/99)(9/98) = .0249
P)2 right-handers) = 3(90/100)(89/99)(10/98) = .2478
P(3 right-handers) = (90/100)(89/99)(88/98) = .7265
6.56
a P(win both) = .28

b P(lose both) = .30
c P(win only one) = .12 + .30 = .42
6.57
P(sale) = .04
167
6.58
P(D) = .02 + .018 = .038
6.59
P(Same party affiliation) = P(DD) + P(RR) + P(OO) = .1936 + .1369 + .0361.3666
168
6.60
Diversity index = .12 + .04 + .12 + .0075 + .04 + .0075 = .335
6.61
P(heart attack) = .0504 + .0792 = .1296
169
6.62
P(pass) = .228 + .243 + .227 = .698
6.63
P(good ) = .3132 + .0416 = .3548
170
6.64
P(myopic) = .1008 + .1512 = .2520
6.65
P(does not have to be discarded) = .1848 + .78 = .9648
6.66 Let A = mutual fund outperforms the market in the first year
B = mutual outperforms the market in the second year
P(A and B) = P(A)P(B | A) = (.15)(.22) = .033
6.67 Let A = DJIA increase and B = NASDAQ increase

P(A) = .60 and P(B | A) = .77
P(A and B) = P(A)P(B | A) = (.60)(.77) = .462
6.68 Define the events:

M: The main control will fail.
B1: The first backup will fail.
B2: The second backup will fail
171
The probability that the plane will crash is
P(M and B1 and B2) = [P(M)][ P(B1)][ P(B2)]
= (.0001) (.01) (.01)
= .00000001
We have assumed that the 3 systems will fail independently of one another.
6.69 P( wireless Web user uses it primarily for e-mail) = .69

P(3 wireless Web users use it primarily for e-mail) = (.69)(.69)(.69) = .3285
6.70
P(Increase) = .05 + .5625 = .6125
6.71 P(A and B) = .36, P(B) = .36 + .07 = .43

P(A and B) .36
P(A | B) = = = .837
P(B) .43
6.72 P(A and B) = .32, P(AC and B) = .14, P(B) = .46, P(BC) = .54
P(A and B) .32
a P(A | B) = = = .696
P(B) .46
P(A C and B) .14

b P(AC | B) = = = .304
P(B) .46
P(A and B C ) .48

c P(A and BC) = .48; P(A | BC ) = = = .889
C .54
P( B )
P(A C and B C ) .06

d P(AC and BC) = .06; P(AC | BC) = = = .111
C .54
P( B )
172
6.73
P(B) = .4940 + .0115 = .5055

P(A and B) .4940
P(A | B) = = = .9773
P(B) .5055
P(F and D) .020

6.74 P(F | D) = = = .526
P(D) .038
6.75 Define events: A = crash with fatality, B = BAC is greater than .09)
P(A) = .01, P(B | A) = .084, P(B) = .12
P(A and B) = (.01)(.084) = .00084
P(A and B) .00084
P(A | B) = = = .007
P(B) .12
P(CFA I and passed) .228

6.76 P(CFA I | passed) = = = .327
P(passed) .698
6.77 Define events: A = heart attack, B = periodontal disease

P(A) = .10, P(B | A) = .85, P(B | AC ) = .29
173
P(B ) = .085 + .261 = .346
P(A and B) .085
P(A | B) = = = .246
P(B) .346
6.78 P(A) = .40, P(B | A) = .85, P(B | AC ) = .29
P(B ) = .34 + .174 = .514

P(A and B) .34
P(A | B) = = = .661
P(B) .514
6.79 Define events: A = smoke, B1 = did not finish high school, B 2 = high school graduate, B 3 =
some college, no degree, B 4 = completed a degree
P(A | B1 ) = .40, P(A | B 2 ) = .34, P(A | B 3 ) = .24, P(A | B 4 ) = .14
From Exercise 6.45: P( B1 ) = .1055, P( B 2 ) = .3236, P( B 3 ) = .1847, P( B 4 ) = .3862
174
P(A) = .0422 + .1100 + .0443 + .0541 = .2506
P( B 4 | A) = .0541/.2506 = .2159
6.80 Define events: A, B, C = airlines A, B, and C, D = on time

P(A) = .50, P(B) = .30, P(C) = .20, P(D | A) = .80, P(D | B) = .65, P(D | C) = .40
P(D) = .40 + .195 + .08 = .675

P(A and D) .40
P(A | D) = = = .593
P(D) .675
6.81 Define events: A = win series, B = win first game
175
P(A) = .60, P(B | A) = .70, P(B | AC ) =.25
P(BC ) = .18 + .30 = .48
P(A and B C ) .18

P(A | BC ) = = = .375
C .48
P( B )
6.82
P(PT) = .28 + .052 = .332

P(R and PT) .28
P(R | PT) = = = .843
P(PT) .332
176
6.83
P(PT) = .0046 + .0269 = .0315

P(H and PT) .0046
P(H | PT) = = = .1460
P(PT) .0315
6.84 Sensitivity = P(PT | H) = .920
Specificity = P(NT | H C ) = .973

Positive predictive value = P(H | PT) = .1460
P(H C and NT) .9681 .9681

Negative predictive value = P H C | NT) = = = = .9996
P( NT) .0004 + .9681 .9685
6.85
177
P(PT) = .0164 + .6233 = .6397
P(NT) = .0036 + .3567 = .3603
P(C and PT) .0164
P(C | PT) = = = .0256
P(PT) .6397
P(C and NT) .0036

P(C | NT) = = = .0010
P( NT) .3603
6.86 a P(Marketing A) = .053 + .237 = .290

b P(Marketing A | Statistics not A) =
P(Marketing A and Statistics not A) .23 .237
= = = .290
P(Statistics not A) .237 + .580 .817
c Yes, the probabilities in Parts a and b are the same.
6.87 Define events: A = win contract A and B = win contract B
a P(A and B) = .12

b P(A and BC) + P(AC and B ) = .18 + .14 = .32
c P(A and B) + P(A and BC ) + P(AC and B ) = .12 + .18 + .14 = .44
6.88 a P(second) = .05 + .14 = .19

P(successful and − 8 or less) .15 .15
b P(successful | –8 or less) = = = = .517
P(−8 or less) .15 + .14 .29
c No, because P(successful) = .66 + .15 = .81, which is not equal to P(successful | –8 or less) .
178
6.89 Define events: A = woman, B = drug is effective
P(B) = .528 + .221 = .749
P(A C and B) .221

6.90 P(AC | B) = = = .295
P(B) .749
6.91 P(Idle roughly)
= P(at least one spark plug malfunctions) = 1– P(all function) = 1 – (.90 4 ) = 1-.6561 = .3439
6.92
P(no sale) = .65 + .175 = .825
6.93 a P(pass) = .86 + .03 = .89

P(pass and miss 5 or more classes) .03 .03
b P(pass | miss 5 or more classes) = = = = .250
P(miss 5 or more classes) .09 + .03 .12
P(pass and miss less than 5 classes) .86 .86

c P(pass | miss less than 5 classes) = = = = .977 .
P(miss less than 5 classes) .86 + .02 .88
d No since P(pass) ≠ P(pass | miss 5 or more classes)
179
6.94 Define events: R = reoffend, D = detained
a P(D) = P(R and D) + P(R C and D) = .1107 + .2263 = .3370

P(R and D) .1107
P(R| D) = = = .3285
P(D) .3370
b P(D C ) = P(R and D C ) + P(R C and D C ) = .1593 + .5037 = .6630
P(R and D C ) .1593

P(R| D C ) = = = .2403
C .6630
P( D )
6.95 a P(excellent) = .27 + .22 = .49

P(man and excellent) .22 .22
b P(excellent | man) = = = = .44
P(man) .22 + .10 + .12 + .06 .50
P(man and excellent) .22 .22

c P(man | excellent) = = = = .449
P(excellent) .27 + .22 .49
d No, since P(excellent) ≠ P(excellent | man)
180
6.96
P(R) = .0176 + .5888 = .6064

P(S and R ) .5888
P(S | R) = = = .9710
P( R ) .6064
6.97 Define events: A1 = Low-income earner, A 2 = medium-income earner, A 3 = high-income
earner, B = die of a heart attack, BC survive a heart attack
P(BC ) = .1848 + .4459 + .2790 = .9097
P(A1 and B C ) .1848

P( A1 | BC ) = = = .2031
C .9097
P( B )
181
6.98 Define the events: A1 = envelope containing two Maui brochures is selected, A 2 = envelope
containing two Oahu brochures is selected, A 3 = envelope containing one Maui and one Oahu
brochures is selected. B = a Maui brochure is removed from the selected envelope.
P(B) = 1/3 + 0 + 1/6 = 1/2

P(A1 and B) 1 / 3
P( A1 | B) = = = 2/3
P(B) 1/ 2
6.99 Define events: A = purchase extended warranty, B = regular price

P(A and B) .21 .21
a P(A | B) = = = = .2692
P(B) .21 + .57 .78
b P(A) = .21 + .14 = .35
c No, because P(A) ≠ P(A | B)
6.100 Define events: A = company fail, B = predict bankruptcy
P(B) = .068 + .2392 = .3072
182
P(A and B) .068
P(A | B) = = = .2214
P(B) .3072
6.101 Define events: A = job security is an important issue, B = pension benefits is an important
issue
P(A) = .74, P(B) = .65, P(A | B) = .60
a P(A and B) = P(B)P(A | B) = (.65)(.60) = .39
b P(A or B) = .74 + .65 – .39 = 1
6.102 Probabilities of outcomes: P(HH) = .25, P(HT) = .25, P(TH) = .25, P(TT) = .25
P(TT | HH is not possible) = .25/(.25 + .25 + .25) = .333
6.103 P(T) = .5
Case 6.1
1. P(Curtain A) = 1/3, P(Curtain B) = 1/3
2. P(Curtain A) = 1/3, P(Curtain B) = 2/3
Switch to Curtain B and double your probability of winning the car.
Case 6.2
Probability Bases Probability Joint
Outcome of outcome Occupied Outs of scoring Probability
1 .75 2nd 1 .42 .3150

2 .10 1st 1 .26 .0260
3 .10 none 2 .07 .0070
4 .05 1st and 2nd 0 .59 .0295
P(scoring) = .3775
Because the probability of scoring with a runner on first base with no outs (.39) is greater than the
probability of scoring after bunting (.3775) you should not bunt.
Case 6.3
0 outs:
Probability of scoring any runs from first base = .39
Probability of scoring from second base = probability of successful steal × probability of scoring
any runs from second base = (.68)(.57) = .3876
Decision: Do not attempt to steal.
183
1 out:
any runs from second base = (.68) × (.42) = .2856
Decision: Attempt to steal.
2 outs:
any runs from second base = (.68) × (.24) = .1632
Decision: Attempt to steal.
Case 6.4
Age 25: P(D) = 1/1,300

P(D and PT) = (1/1,300)(.624) = .00048
P(D and NT) = (1/1,300)(.376) = .00029
P( D C and PT) = (1,299/1,300)(.04) = .03997
P( D C and NT) = (1,299/1,300)(.96) = .95926

P(PT) = .00048 + .03997 = .04045
P(NT) = .00029 + .95926 = .95955
P(D | PT) = .00048/.04045 = .01187
P(D | NT) = .00029/.95955 = .00030
184
Age 30: P(D) = 1/900
P(D and PT) = (1/900)(.710) = .00079
P(D and NT) = (1/900)(.290) = .00032
P( D C and PT) = (899/900)(.082) = .08190

P( D C and NT) = (899/900)(.918) = .91698
P(PT) = .00079 + .08190 = .08269
P(NT) = .00032 + .91698 = .91730
P(D | PT) = .00079/.08269 = .00955
P(D | NT) = .00032/.91730 = .00035
Age 35: P(D) = 1/350

P(D and PT) = (1/350)(.731) = .00209
P(D and NT) = (1/350)(.269) = .00077
P( D C and PT) = (349/350)(.178) = .17749
P( D C and NT) = (349/350)(.822) = .81965

P(PT) = .00209 + .17749 = .17958
P(NT) = .00077 + .81965 = .82042
P(D | PT) = .00209/.17958 = .01163
P(D | NT) = .00077/.82042 = .00094
Age 40: P(D) = 1/100

P(D and PT) = (1/100)(.971) = .00971
P(D and NT) = (1/100)(.029) = .00029
P( D C and PT) = (99/100)(.343) = .33957
P( D C and NT) = (99/100)(.657) = .65043

P(PT) = .00971 + .33957 = .34928
P(NT) = .00029 + .65043 = .65072
P(D | PT) = .00971/.34928 = .02780
P(D | NT) = .00029/.65072 = .00045
Age 45: P(D) = 1/25

P(D and PT) = (1/25)(.971) = .03884
P(D and NT) = (1/25)(.029) = .00116
P( D C and PT) = (24/25)(.343) = .32928
P( D C and NT) = (24/25)(.657) = .63072

P(PT) = .03884 + .32928 = .36812
185
P(NT) = .00116 + .63072 = .63188
P(D | PT) = .03884/.36812 = .10551
P(D | NT) = .00116/.63188 = .00184
Age 49: P(D) = 1/12

P(D and PT) = (1/12)(.971) = .08092
P(D and NT) = (1/12)(.029) = .00242
P( D C and PT) = (11/12)(.343) = .31442
P( D C and NT) = (11/12)(.657) = .60255

P(PT) = .08092 + .31442 = .39533
P(NT) = .00242 + .60255 = .60467
P(D | PT) = .08092/.39533 = .20468
P(D | NT) = .00242/.60467 = .00400
Case 6.5
The probability that 23 people have different birthdays is .4927. The probability that at least two
people have the same birthday is 1 − .4927 = .5073.
186
Chapter 7
7.1 a 0, 1, 2, …
b Yes, we can identify the first value (0), the second (1), and so on.
c It is finite, because the number of cars is finite.
d The variable is discrete because it is countable.
7.2 a any value between 0 and several hundred miles

b No, because we cannot identify the second value or any other value larger than 0.
c No, uncountable means infinite.
d The variable is continuous.
7.3 a The values in cents are 0 ,1 ,2, …

b Yes, because we can identify the first ,second, etc.
c Yes, it is finite because students cannot earn an infinite amount of money.
d Technically, the variable is discrete.
7.4 a 0, 1, 2, …, 100
b Yes.
c Yes, there are 101 values.
d The variable is discrete because it is countable.
7.5 a No the sum of probabilities is not equal to 1.

b Yes, because the probabilities lie between 0 and 1 and sum to 1.
c No, because the probabilities do not sum to 1.
7.6 P(x) = 1/6 for x = 1, 2, 3, 4, 5, and 6
7.7 a x P(x)
0 1218/101,501 = .012
1 32,379/101,501 = .319
2 37,961/101,501 = .374
3 19,387/101,501 = .191
4 7714/101,501 = .076
5 2842/101,501 = .028
b (i) P(X ≤ 2) = P(0) + P(1) + P(2) = .012 + .319 + .374 = .705
(ii) P(X > 2) = P(3) + P(4) + P(5) = .191 + .076 + .028 = .295
(iii) P(X ≥ 4) = P(4) + P(5) = .076 + .028 = .104
187
7.8 a P(2 ≤ X ≤ 5) = P(2) + P(3) + P(4) + P(5) = .310 + .340 + .220 + .080 = .950
P(X > 5) = P(6) + P(7) = .019 + .001 = .020
P(X < 4) = P(0) + P(1) + P(2) + P(3) = .005 + .025 + .310 + .340 = .680
b. b. E(X) = ∑ xP(x) = 0(.005) + 1(.025) + 2(.310) + 4(.340) +5(.080) + 6(.019) + 7(.001) =

3.066
2 2 2
c. σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–3.066) (.005) + (1–3.066) (.025) + (2–3.066) (.310)
2 2 2 2
+ (3–3.066) (.340) + (4–3.066) (.220) + (5–3.066) (.080) + (6–3.066) (.019)
2
+ (7–3.066) (.001) = 1.178
σ= σ 2 = 1.178 = 1.085
7.9 P(0) = P(1) = P(2) = . . . = P(10) = 1/11 = .091
7.10 a P(X > 0) = P(2) + P(6) + P(8) = .3 + .4 + .1 = .8

b P(X ≥ 1) = P(2) + P(6) + P(8) = .3 + .4 + .1 = .8
c P(X ≥ 2) = P(2) + P(6) + P(8) = .3 + .4 + .1 = .8
d P(2 ≤ X ≤ 5) = P(2) = .3
7.11a P(3 ≤ X ≤ 6) = P(3) + P(4) + P(5) + P(6) = .04 + .28+ .42 + .21 = .95
b. P(X > 6) = P(X ≥ 7) = P(7) + P(8) = .02 + .02 = .04
c. P(X < 3) = P(X ≤ 2) = P(0) + P(1) + P(2) = 0 + 0 + .01 = .01
7.12 P(Losing 6 in a row) = .5 6 = .0156
7.13 a P(X < 2) = P(0) + P(1) = .05 + .43 = .48

b P(X > 1) = P(2) + P(3) = .31 + .21 = .52
188
7.14
a P(HH) = .25
b P(HT) = .25
c P(TH) = .25
d P(TT) = .25
7.15 a P(0 heads) = P(TT) = .25

b P(1 head) = P(HT) + P(TH) = .25 + .25 = .50
c P(2 heads) = P(HH) = .25
d P(at least 1 head) = P(1 head) + P(2 heads) = .50 + .25 = .75
7.16
189
7.17 a P(2 heads) = P(HHT) + P(HTH) + P(THH) = .125 + .125 + .125 = .375
b P(1 heads) = P(HTT) + P(THT) = P(TTH) = .125 + .125 + .125 = .375
c P(at least 1 head) = P(1 head) + P(2 heads) + P(3 heads) = .375 + .375 + .125 = .875
d P(at least 2 heads) = P(2 heads) + P(3 heads) = .375 + .125 = .500
7.18a. µ = E(X) = ∑ xP(x) = –2(.59) +5(.15) + 7(.25) +8(.01) = 1.40

2 2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (–2–1.4) (.59) + (5–1.4) (.15) + (7–1.4) (.25) + (8–1.4) (.01)
= 17.04
b. x –2 5 7 8
y –10 25 35 40
P(y) .59 .15 .25 .01
∑ yP( y) = –10(.59) + 25(.15) + 35(.25) + 40(.01) = 7.00

c. E(Y) =
2 2 2 2
V(Y) = ∑ ( y − µ) P( y) = (–10–7.00) (.59) + (25–7.00) (.15) + (35–7.00) (.25)
2
+ (40–7.00) (.01) = 426.00
d. E(Y) = E(5X) = 5E(X) = 5(1.4) = 7.00
V(Y) = V(5X) = 5 2 V(X) = 25(17.04) = 426.00.
7.19a µ = E(X) = ∑ xP(x) = 0(.4) + 1(.3) + 2(.2) + 3(.1) = 1.0

2 2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–1.0) (.4) + (1–1.0) (.3) + (2–1.0) (.2) + (3–1.0) (.1)
= 1.0
σ= σ 2 = 1.0 = 1.0
b. x 0 1 2 3
y 2 5 8 11
P(y) .4 .3 .2 .1
c. E(Y) = ∑ yP( y) = 2(.4) + 5(.3) + 8(.2) + 11(.1) = 5.0

2 2 2 2
σ2 2
=V(Y) = ∑ ( y − µ) P( y) = (2 – 5) (.4) + (5 – 5) (.3) + (8 – 5) (.2) + (11 – 5) (.1) = 9.0
σ= σ 2 = 9.0 = 3.0
d. E(Y) = E(3X + 2) = 3E(X) + 2 = 3(1) + 2 = 5.0
σ 2 = V(Y) = V(3X + 2) = V(3X) = 3 2 V(X) = 9(1) = 9.0.
σ= σ 2 = 9.0 = 3.0
The parameters are identical.
190
7.20a. P(X ≥ 2) = P(2) + P(3) = .4 + .2 = .6
b. µ = E(X) = ∑ xP(x) = 0(.1) + 1(.3) + 2(.4) + 3(.2) = 1.7

2 2 2 2
σ2 2
= V(X) = ∑ ( x − µ) P( x ) = (0–1.7) (.1) + (1–1.7) (.3) + (2–1.7) (.4) + (3–1.7) (.2) = .81
7.21 E(Profit) = E(5X) = 3E(X) = 3(1.7) = 5.1
V(Profit) = V(3X) = 3 2 V(X) = 9(.81) = 7.29
7.22 a P(X > 4) = P(5) + P(6) + P(7) = .20 + .10 + .10 = .40
b P(X ≥ 2) = 1– P(X ≤ 1) = 1 – P(1) = 1 – .05 = .95
7.23 µ = E(X) = ∑ xP(x) = 1(.05) + 2(.15) + 3(.15) + 4(.25) + 5(.20) + 6(.10) + 7(.10) = 4.1
2 2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (1–4.1) (.05) + (2–4.1) (.15) + (3–4.1) (.15) + (4–4.1) (.25)
2 2 2
+ (5–4.1) (.20) + (6–4.1) (.10) + (7–4.1) (.10) = 2.69
7.24 Y = .25X; E(Y) = .25E(X) = .25(4.1) = 1.025

2
V(Y) = V(.25X) = (.25) (2.69) = .168
7.25 a. x 1 2 3 4 5 6 7
y .25 .50 .75 1.00 1.25 1.50 1.75
P(y) .05 .15 .15 .25 .20 .10 .10
b. E(Y) = ∑ yP( y) = .25(.05) + .50(.15) + .75(.15) +1.00(.25) + 1.25(.20) + 1.50(.10) + 1.75(.10)

= 1.025
2 2 2 2
V(Y) = ∑ ( y − µ) P( y) = (.25–1.025) (.05) + (.50–1.025) (.15) + (.75–1.025) (.15)
2 2 2
+ (1.00–1.025) (.25) + (1.25–1.025) (.20) + (1.50–1.025) (.10) + (1.75–
2
1.1025) (.10) = .168
c. The answers are identical.
7.26 a P(4) = .06

b P(8) = 0
c P(0) = .35
d P(X ≥ 1) = 1 – P(0) = 1 – .35 = .65
7.27 a P(X ≥ 20) = P(20) + P(25) + P(30) + P(40) + P(50) + P(75) + P(100)
191
= .08 + .05 + .04 + .04 + .03 + .03 + .01 = .28
b P(X = 60) = 0
c P(X > 50) = P(75) + P(100) = .03 + .01 = .04
d P(X > 100) = 0
7.28 a P(X = 3) = P(3) = .21

b P(X ≥ 5) = P(5) + P(6) + P(7) + P(8) = .12 + .08 + .06 + .05 = .31
c P(5 ≤ X ≤ 7) = P(5) + P(6) + P(7) = .12 + .08 + .06 = .26
7.29 a P(X > 1) = P(2) + P(3) + P(4) = .17 + .06 + .01 = .24
b P(X = 0) = .45
c P(1 ≤ X ≤ 3) = P(1) + P(2) + P(3) = .31 + .17 + .06 = .54
7.30 µ = E(X) = ∑ xP(x) = 0(.04) + 1(.19) + 2(.22) + 3(.28) + 4(.12) + 5(.09) + 6(.06) = 2.76
2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (1–2.76) (.04) + (2–2.76) (.19) + (3–2.76) (.28)
2 2 2
+ (4–2.76) (.12) + (5–2.76) (.09) + (6–2.76) (.06) = 2.302
σ= σ 2 = 2.302 = 1.517
7.31 Y = 10X; E(Y) = E(10X) = 10E(X) = 10(2.76) = 27.6

2
V(Y) = V(10X) = 10 V(X) =100(2.302) = 230.2
σ= σ 2 = 230.2 = 15.17
7.32 µ = E(X) = ∑ xP(x) = 1(.24) + 2(.18) + 3(.13) + 4(.10) + 5(.07) + 6(.04) + 7(.04) + 8(.20) =
3.86
2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (1–3.86) (.24) + (2–3.86) (.18) + (3–3.86) (.13) + (4–
2
3.86) (.10)
2 2 2 2
+ (5–3.86) (.07) +(6–3.86) (.04) + (7–3.86) (.04) + (8–3.86) (.20) = 6.78
σ= σ 2 = 6.78 = 2.60
7.33 Revenue = 2.50X; E(Revenue) = E(2.50X) = 2.50E(X) = 2.50(3.86) = 9.65
V(Revenue) = V(2.50X) = 2.50 2 (V(X) = 6.25(6.78) = 42.38
σ= σ 2 = 42.38 = 6.51
192
7.34 E(Value of coin) = 400(.40) + 900(.30) + 100(.30) = 460. Take the $500.
7.35 µ = E(X) = ∑ xP(x) = 0(.10) + 1(.20) + 2(.25) + 3(.25) + 4(.20) = 2.25

2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–2.25) (.10) + (1–2.25) (.20) + (2–2.25) (.25) + (3–
2
2.25) (.13)
2
+ (4–2.25) (.20) = 1.59
σ= σ 2 = 1.59 = 1.26
7.36 E(damage costs) = .01(400) + .02(200) + .10(100) + .87(0) = 18. The owner should pay up to
$18 for the device.
7.37 E(X) = ∑ xP(x) = 1,000,000(1/10,000,000) + 200,000(1/1,000,000) + 50,000(1/500,000)

+ 10,000(1/50,000) + 1,000(1/10,000) = .1 + .2 + .1 + .2 + .1 = .7
Expected payoff = 70 cents.
7.38 µ = E(X) = ∑ xP(x) = 1(.05) + 2(.12) + 3(.20) + 4(.30) + 5(.15) + 6(.10) + 7 (.08) = 4.00
2 2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (1–4.0) (.05) + (2–4.0) (.12) + (3–4.0) (.20) + (4–4.0) (.30)
2 2 2
+ (5–4.0) (.15) +(6–4.0) (.10) + (7–4.0) (.08) = 2.40
7.39 Y = .25X; E(Y) = E(.25X) = .25E(X) = .25(4.0) = 1.0

2
V(Y) = V(.25X) = (.25) V(X) =.0625(2.40) = .15
7.40 µ = E(X) = ∑ xP(x) = 0(.10) + 1(.25) + 2(.40) + 3(.20) + 4(.05) = 1.85

7.41 Profit = 4X; Expected profit = E(4X) = 4E(X) = 4(1.85) = $7.40
7.42 Breakeven point = 15,000/(7.40 – 3.00) = 3,409
7.43 a x P(x)
1 .6
2 .4
193
b y P(y)
1 .6
2 .4
c µ = E(X) = ∑ xP(x) = 1(.6) + 2(.4) = 1.4

2 2
σ2 2
= V(X) = ∑ ( x − µ) P( x ) = (1–1.4) (.6) + (2–1.4) (.4) = .24
d µ = 1.4, σ 2 = .24
7.44 a ∑ ∑ xyP(x, y) = (1)(1)(.5) + (1)(2)(.1) + (2)(1)(.1) + (2)(2)(.3) = 2.1

all x all y
COV(X, Y) = ∑ ∑ xyP(x, y) – µ x µ y = 2.1 – (1.4)(1.4) = .14

all x all y
σ x = σ 2x = .24 = .49, σ y = σ 2y = .24 = .49
COV (X, Y) .14

ρ= = = .58
σxσy (.49)(.49)
7.45 E(X + Y) = E(X) + E(Y) = 1.4 + 1.4 = 2.8

V(X + Y) = V(X) + V(Y) + 2COV(X, Y) = .24 + .24 + 2(.14) = .76
7.46 a x+y P(x + y)

2 .5
3 .2
4 .3
b µ x+ y = E(X+Y) = ∑ (x + y)P(x + y) = 2(.5) + 3(.2) + 4(.3) = 2.8

2 2 2
σ 2x+ y = V(X+Y) = ∑ [(x + y) − µ 2
] P( x + y) = (2–2.8) (.5) + (3–2.8)
x+y (.2) + (4–2.8) (.3) =
.76
c Yes
7.47 a x P(x)
1 .4
2 .6
b y P(y)
1 .7
2 .3
c µ = E(X) = ∑ xP(x) = 1(.4) + 2(.6) = 1.6

194
2 2
σ 2 = V(X) = 2
∑ (x − µ) P(x) = (1–1.6) (.4) + (2–1.6) (.6) = .24
d µ = E(Y) = ∑ yP( y) = 1(.7) + 2(.3) = 1.3
2 2 2 2
σ = V(Y) = ∑ ( y − µ) P( y) = (1–1.3) (.7) + (2–1.3) (.3) = .21
7.48 a ∑ ∑ xyP(x, y) = (1)(1)(.28) + (1)(2)(.12) + (2)(1)(.42) + (2)(2)(.18) = 2.08

all x all y
COV(X, Y) = ∑ ∑ xyP(x, y) – µ x µ y = 2.08 – (1.6)(1.3) = 0

all x all y
σ x = σ 2x = .24 = .49, σ y = σ 2y = .21 = .46
COV (X, Y) 0
ρ= = =0
σxσy (.49)(.46)
7.49 E(X + Y) = E(X) + E(Y) = 1.6 + 1.3 = 2.9

V(X + Y) = V(X) + V(Y) + 2COV(X, Y) = .24 + .21 + 2(0) = .45
7.50 a x + y P(x + y)
2 .28
3 .54
4 .18
b µ x+ y = E(X+Y) = ∑ (x + y)P(x + y) = 2(.28) + 3(.54) + 4(.18) = 2.9

2 2 2
σ 2x+ y = V(X+Y) = ∑ [(x + y) − µ 2
] P( x + y) = (2–2.9) (.28) + (3–2.9)
x+y (.54) + (4–2.9) (.18)
= .45
c Yes
7.51 a x P(x) y P(y)

1 .7 1 .6
2 .2 2 .4
3 .1
∑ xP(x) = 1(.7) + 2(.2) + 3(.1) = 1.4

b µ x = E(X) =
2 2 2 2 2
σ = V(X) = ∑ ( x − µ) P( x ) = (1–1.4) (.7) + (2–1.4) (.2) + (3–1.4) (.1) = .44
µ = E(Y) = ∑ yP( y) = 1(.6) + 2(.4) = 1.4

y
2 2 2 2
σ = V(Y) = ∑ ( y − µ) P( y) = (1–1.4) (.6) + (2–1.4) (.4) = .24
195
∑ ∑ xyP(x, y) = (1)(1)(.42) + (1)(2)(.28) + (2)(1)(.12) + (2)(2)(.08) + (3)(1)(.06) +
all x all y
(3)(2)(.04) = 1.96
COV(X, Y) = ∑ ∑ xyP(x, y) – µ x µ y = 1.94 – (1.4)(1.4) = 0
all x all y
σ x = σ 2x = .44 = .66, σ y = σ 2y = .24 = .49
COV (X, Y) 0
ρ= = =0
σxσy (.66)(.49)
c x+y P(x + y)
2 .42
3 .40
4 .14
5 .04
7.52 x
y 0 1 2
1 .42 .21 .07
2 .18 .09 .03
7.53 x
y 0 1
1 .04 .16
2 .08 .32
3 .08 .32
7.54 a Refrigerators, x P(x)

0 .22
1 .49
2 .29
b Stoves, y P(y)
0 .34
1 .39
2 .27
c µ x = E(X) =∑ xP(x) = 0(.22) + 1(.49) + 2(.29) = 1.07

2 2 2
σ2 2
= V(X) = ∑ ( x − µ) P( x ) = (0–1.07) (.22) + (1–1.07) (.49) + (2–1.07) (.29) = .505
196
∑ yP( y) = 0(.34) + 1(.39) + 2(.27) = .93
d µ y = E(Y) =
2 2 2 2 2
σ = V(Y) = ∑ ( y − µ) P( y) = (0–.93) (.34) + (1–.93) (.39) + (2–.93) (.27) = .605
e ∑ ∑ xyP( x, y) = (0)(0)(.08) + (0)(1)(.09) + (0)(2)(.05) + (1)(0)(.14) + (1)(1)(.17)

all x all y
+ (1)(2)(.18) + (2)(0)(.12) + (2)(1)(.13) + (2)(2)(.04) = .95
COV(X, Y) = ∑ ∑ xyP(x, y) – µ x µ y = .95 – (1.07)(.93) = –.045

all x all y
σ x = σ 2x = .505 = .711, σ y = σ 2y = .605 = .778
COV (X, Y) −.045

ρ= = = –.081
σxσy (.711)(.778)
7.55 a Bottles, x P(x)

0 .72
1 .28
b Cartons, y P(y)
0 .81
1 .19
∑ xP(x) = 0(.72) + 1(.28) = .28

c µ x = E(X) =
2 2 2 2
σ = V(X) = ∑ ( x − µ) P( x ) = (0–.28) (.72) + (1–.28) (.28) = .202
d µ = E(Y) = ∑ yP( y) = 0(.81) + 1(.19) = .19

y
2 2 2 2
σ = V(Y) = ∑ ( y − µ) P( y) = (0–.19) (.81) + (1–.19) (.19) = .154
e ∑ ∑ xyP( x, y) = (0)(0)(.63) + (0)(1)(.09) + (1)(0)(.18) + (1)(1)(.10) = .100

all x all y
COV(X, Y) = ∑ ∑ xyP(x, y) – µ x µ y = .100 – (.28)(.19) = .0468

all _ x all _ y
σ x = σ 2x = .202 = .449, σ y = σ 2y = .154 = .392
COV (X, Y) .0468

ρ= = = .266
σxσy (.449)(.392)
7.56 a P(X = 1 | Y = 0) = P(X =1 and Y = 0)/P(Y = 0) = .14/.34 = .412

b P(Y = 0 | X = 1) = P(X =1 and Y = 0)/P(X = 1) = .14/.49 = .286
c P(X = 2 | Y = 2) = P(X =2 and Y = 2)/P(Y = 2) = .04/.27 = .148
197
(∑ X ) = ∑ E(X ) = 18 + 12 + 27 + 8 = 65
7.57 E i i
V (∑ X ) = ∑ V (X ) = 8 + 5 + 6 + 2 = 21
i i
(∑ X ) = ∑ E(X ) = 35 + 20 + 20 + 50 + 20 = 145
7.58 E i i
V (∑ X ) = ∑ V (X ) = 8 + 5 + 4 + 12 + 2 = 31
i i
(∑ X ) = ∑ E(X ) = 8 + 14 + 5 + 3 + 30 + 30 + 10 = 100
7.59 E i i
V (∑ X ) = ∑ V (X ) = 2 + 5 + 1 + 1 + 8 +10 + 3 = 30
i i
(∑ X ) = ∑ E(X ) = 10 + 3 + 30 + 5 + 100 + 20 = 168

7.60 E i i
V (∑ X ) = ∑ V (X ) = 9 + 0 + 100 + 1 + 400 + 64 = 574

i i
7.61 The expected value does not change. The standard deviation decreases.
7.62 E(Rp) = w1E(R1) + w2E(R2) = (.30)(.12) + (.70)(.25) = .2110
a. V(Rp) = w12 σ12 + w22 σ 22 + 2 w1 w2 ρ σ1 σ 2

= (.30)2 (.02)2 + (.70)2 (.152 ) + 2(.30)(.70)(.5)(.02)(.15) = .0117
σ R p = .0117 = .1081
b. V(Rp) = w12 σ12 + w22 σ 22 + 2 w1 w2 ρ σ1 σ 2

= (.30) 2 (.02) 2 + (.70) 2 (.15 2 ) + 2(.30)(.70)(.2)(.02)(.15) = .0113
σ R p = .0113 = .1064
c. V(Rp) = w12 σ12 + w22 σ 22 + 2 w1 w2 ρ σ1 σ 2

= (.30) 2 (.02) 2 + (.70) 2 (.15 2 ) + 2(.30)(.70)(0)(.02)(.15) = .0111
σ R p = .0111 = .1052
7.63 a She should choose stock 2 because its expected value is higher.
b. She should choose stock 1 because its standard deviation is smaller.
7.64 E(Rp) = w1E(R1) + w2E(R2) = (.60)(.09) + (.40)(.13) = .1060
V(Rp) = w12 σ12 + w22 σ 22 + 2 w1 w2 ρ σ1 σ 2
198
= (.60) 2 (.15) 2 + (.40) 2 (.212 ) + 2(.60)(.40)(.4)(.15)(.21) = .0212
σ R p = .0212 = .1456
7.65 E(Rp) = w1E(R1) + w2E(R2) = (.30)(.09) + (.70)(.13) = .1180
V(Rp) = w12 σ12 + w22 σ 22 + 2 w1 w2 ρ σ1 σ 2

= (.30) 2 (.15) 2 + (.70) 2 (.212 ) + 2(.30)(.70)(.4)(.15)(.21) = .0289
σ R p = .0289 = .1700
The statistics used in Exercises 7.66 to 7.78 were computed by Excel. The variances were taken
from the variance-covariance matrix. As a result they are the population parameters. To convert
to statistics multiply the variance of the portfolio returns by n/(n–1).
7.66 a Stock Mean Variance

AT&T .00717 .00318
Aetna .00627 .01147
Cigna .01327 .01550
Coca-Cola .00881 .00239
Disney .00562 .00442
Ford .01555 .05104
McDonald’s .01478 .00277
b
AT&T Aetna Cigna Coca Cola Disney Ford McDonalds
AT&T 0.00318
Aetna 0.00111 0.01147
Cigna 0.00162 0.01071 0.01550
Coca Cola 0.00105 0.00154 0.00173 0.00239
Disney 0.00120 0.00308 0.00389 0.00144 0.00442
Ford 0.00332 0.00168 0.00982 0.00284 0.00741 0.05104
McDonalds 0.00103 0.00196 0.00183 0.00148 0.00148 0.00157 0.00277
7.67The stocks with the largest mean returns are Ford (mean = .01555) and McDonald’s (mean =
.01478)
199
Portfolio of 2 Stocks
Ford McDonalds
Variance-Covariance Matrix Ford 0.05019
McDonalds 0.00154 0.00273
Expected Returns 0.01555 0.01478
Weights 0.5000 0.5000
Portfolio Return
Expected Value 0.01517
Variance 0.01400
Standard Deviation 0.11833
The expected value is .01517 and the standard deviation is .11833.
7.68 The stocks with the smallest variances are Coca-Cola (.00239) and McDonalds (.00277).
Coca Cola McDonalds
Variance-Covariance Matrix Coca Cola 0.00235
McDonalds 0.00145 0.00273
Weights 0.5000 0.5000
Portfolio Return
Variance 0.00200
7.69 The two-stock portfolio with the largest expected value is composed of Ford and
McDonald’s, the two stocks with the highest means. Its expected value is .01517 and its standard
deviation is .11833. The two-stock portfolio with the smallest variance is composed of Coca-Cola
and McDonald’s, the two stocks with both the smallest variances. The expected value is .01180
and the standard deviation is .04469.
200
7.70
AT&T Coca Cola Disney
Variance-Covariance Matrix AT&T 0.00313
Coca Cola 0.00103 0.00235
Disney 0.00118 0.00141 0.00434
Expected Returns 0.00717 0.00881 0.00562
Weights 0.33333 0.33333 0.33333
Portfolio Return
Variance 0.00190
7.71
AT&T Cigna Disney Ford
Cigna 0.001594 0.015245
Disney 0.001178 0.003829 0.004342
Ford 0.003267 0.009658 0.007288 0.050194
Expected Returns 0.00717 0.01327 0.00562 0.01555
Weights 0.30000 0.20000 0.40000 0.10000
Portfolio Return
Variance 0.00434
201
7.72
AT&T Cigna Disney Ford
Cigna 0.001594 0.015245
Disney 0.001178 0.003829 0.004342
Ford 0.003267 0.009658 0.007288 0.050194
Weights 0.30000 0.10000 0.40000 0.20000
Portfolio Return
Variance 0.00577
7.73a Stock Mean Variance

Barrick Gold .01253 .01194
Bell Canada Enterprises .00590 .00428
Bank of Montreal .00593 .00543
Enbridge .01197 .00217
Fortis .01254 .00328
Methanex .00902 .01515
Research in Motion .02536 .02551
Telus .00436 .00488
Trans Canada pipelines .00741 .00188
b
Barrick BCE BMO Enbridge Fortis Methanex RIM Telus TRP
Barrick 0.01194
BCE -0.00262 0.00428
BMO -0.00104 0.00105 0.00543
Enbridge -0.00004 0.00058 0.00025 0.00217
Fortis 0.00025 0.00002 0.00033 0.00162 0.00328
Methanex 0.00114 0.00089 0.00202 0.00069 0.00020 0.01515
RIM -0.00268 0.00262 0.00380 0.00156 0.00003 0.00688 0.02551
Telus -0.00087 0.00181 -0.00019 0.00079 0.00062 -0.00071 0.00069 0.00488
TRP 0.00070 0.00093 0.00039 0.00140 0.00126 0.00121 0.00062 0.00070 0.00188
7.74 The stocks with the largest means are Fortis (.01254) and Research in Motion (.02536).
202
Fortis RIM
Variance-Covariance Matrix Fortis 0.00322
RIM 0.00003 0.02509
Weights 0.50000 0.50000
Portfolio Return
Variance 0.00709
7.75 The stocks with the smallest variances are Enbridge (.00217) and Trans Canada Pipelines
(.00188).
Enbridge TRP
Variance-Covariance Matrix Enbridge 0.00213
TRP 0.00138 0.00184
Weights 0.50000 0.50000
Portfolio Return
Variance 0.00168
7.76 The two-stock portfolio with the largest expected value is composed of Fortis and Research
in Motion , the two stocks with the highest means. Its expected value is .01895 and its standard
deviation is .08421. The two-stock portfolio with the smallest variance is composed of Enbridge
and Trans Canada Pipelines, the two stocks with the smallest variances. The expect value is
.00969 and the standard deviation is .04103.
203
7.77
BMO Enbridge Fortis
Variance-Covariance Matrix BMO 0.00534
Enbridge 0.00024 0.00213
Fortis 0.00033 0.00159 0.00322
Expected Returns 0.00593 0.01197 0.01254
Weights 0.20000 0.30000 0.50000
Portfolio Return
Variance 0.00178
7.78
Barrick BCE Telus TRP
Variance-Covariance Matrix Barrick 0.01174
BCE -0.00257 0.00421
Telus -0.00085 0.00178 0.00480
TRP 0.00069 0.00091 0.00069 0.00184
Weights 0.50000 0.25000 0.15000 0.10000
Portfolio Return
Variance 0.00282
204
7.79
Barrick BCE Telus TRP
Variance-Covariance Matrix Barrick 0.01174
BCE -0.00257 0.00421
Telus -0.00085 0.00178 0.00480
TRP 0.00069 0.00091 0.00069 0.00184
Weights 0.20000 0.40000 0.20000 0.20000
Portfolio Return
Variance 0.00147
7.80a Stock Mean Variance

Amazon .02834 .02055
Amgen .00150 .00788
Apple .03927 .01442
Cisco Systems .00664 .00628
Google .02497 .01109
Intel .00288 .00651
Microsoft .00618 .00538
Oracle .01203 .00446
Research in Motion .02953 .02990
b. Stocks, Amazon, Apple, Google, and Research in Motion have the largest means.
c. Stocks Cisco, Intel, Microsoft, and Oracle have the smallest variances.
205
7.81
Cisco Intel Microsoft RIM
Variance-Covariance Matrix Cisco 0.00617
Intel 0.00367 0.00640
Microsoft 0.00259 0.00279 0.00529
RIM 0.00741 0.00578 0.00375 0.02940
Weights 0.30000 0.15000 0.25000 0.30000
Portfolio Return
Variance 0.00702
The expected value is .01283 and the variance is .08379.
7.83
Cisco Intel Microsoft RIM
Variance-Covariance Matrix Cisco 0.00617
Intel 0.00367 0.00640
Microsoft 0.00259 0.00279 0.00529
RIM 0.00741 0.00578 0.00375 0.02940
Weights 0.26586 0.02491 0.54737 0.16186
Portfolio Return
Variance 0.00502
The expected value is .01000 and the variance is .00502.
n!
7.84 P(X = x) = p x (1 − p) n −x
x! (n − x )!
10!
a P(X = 3) = (.3) 3 (1 − .3)10−3 = .2668
3! (10 − 3)!
10!
b P(X = 5) = (.3) 5 (1 − .3)10−5 = .1029
5! (10 − 5)!
10!
c P(X = 8) = (.3) 8 (1 − .3)10−8 = .0014
8!(10 − 8)!
206
7.85 a P(X = 3) = P(X≤ 3) – P(X ≤ 2) = .6496 – .3828 = .2668
b P(X = 5) = P(X ≤ 5) – P(X ≤ 4) = .9527 – .8497 = .1030
c P(X = 8) = P(X ≤ 8) – P(X ≤ 7) = .9999 – .9984 = .0015
7.86 a .26683
b .10292
c .00145
n!
7.87 P(X = x) = p x (1 − p) n −x
x! (n − x )!
6!
a P(X = 2) = (.2) 2 (1 − .2) 6−2 = .2458
2! (6 − 2)!
6!
b P(X = 3) = (.2) 3 (1 − .2) 6−3 = .0819
3! (6 − 3)!
6!
c P(X = 5) = (.2) 5 (1 − .2)5 = .0015
5! (6 − 5)!
7.88 a P(X = 2) = P(X≤ 2) – P(X ≤ 1) = .9011 – .6554 = .2457

b P(X = 3) = P(X ≤ 3) – P(X ≤ 2) = .9830 −.9011 = .0819
c P(X = 5) = P(X ≤ 5) – P(X ≤ 4) = .9999 – 9984 = .0015
7.89 a .24576
b .08192
c .00154
7.90 a P(X = 18) = P(X ≤ 18) – P(X ≤ 17) = .6593 – .4882 = .1711
b P(X = 15) = P(X ≤ 15) – P(X ≤ 14) =.1894 – .0978 = .0916
c P(X ≤ 20) = .9095
d P(X ≥ 16) = 1 – P(X ≤ 15) = 1 – .1894 = .8106
7.91 a .17119
b .09164
c .90953
d .81056
7.92 Binomial distribution with p = .25
207
4!
a P(X = 1) = (.25)1(1 − .25)4−1 = .4219
1! (4 − 1)!
b Table 1 with n = 8: p(2) = P(X ≤ 2) – P(X ≤ 1) = .6785 – .3671 = .3114
c Excel: p(3) = .25810
7.93 Table 1 with n = 25 and p = .3: P(X ≤ 10) = .9022
7.94 Table 1 with n = 25 and p = .90

a P(X = 20) = P(X ≤ 20) – P(X ≤ 19) = .0980 – .0334 = .0646
b P(X ≥ 20) = 1 – P(X ≤ 19) = 1 – .0334 = .9666
c P(X ≤ 24) = .9282
d E(X) = np = 25(.90) = 22.5
7.95 Table 1 with n = 25 and p = .75: P(X ≥ 15) = 1 – P(X ≤ 14) = 1 – .0297 = .9703
4!
7.96 P(X = 0) = (.7) 0 (1 − .7) 4−0 = .0081
01!(4 − 0)!
7.97 Table 1 with n = 25 and p = .10

a P(X = 0) = P(X ≤ 0) = .0718
b P(X < 5) = P(X ≤ 4) = .9020
c P(X > 2) = P(X ≥ 3) = 1 – P(X ≤ 2) = 1 – .5371 = .4629
25!
7.98 P(X = 0) = (.08) 0 (1 − .08) 25−0 = .1244
0! (25 − 0)!
7.99 Excel with n = 100 and p = .20: P(X > 25) = P(X ≥ 26) = 1 – P(X ≤ 25) = 1 – .91252 =
.08748
20!
7.100 P(X = 20) = (.75) 20 (1 − .75) 20−20 = .00317
20! (20 − 20)!
7.01a Excel with n = 10 and p = 244/495: P(X ≥ 5) = 1 – P(X ≤ 4) = 1 – .39447 = .60553

b E(X) = np =100(244/495) = 49.29
5!
7.102 a P(X = 2) = (.45) 2 (1 − .45) 5−2 = .3369
2! (5 − 2)!
208
b Excel with n = 25 and p = .45: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .24237 = .75763
7.103 a Table 1 with n = 5 and p = .5: P(X = 2) = P(X ≤ 2) – P(X ≤ 1) = .5 – .1875 = .3125
b: Table 1 with n = 25 and p = .5: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .1148 = .8852
5!
7.104 a P(X = 2) = (.52)2 (1 − .52)5−2 = .2990
2!(5 − 2)!
b Excel with n = 25 and p = .52: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .08033 = .91967
7.105 a Excel with n = 25 and p = 2/38: P(X ≥ 2) = 1 – P(X ≤ 1) = 1 – .61826 = .38174

b Excel with n = 25 and p = 2/38: P(X = 0)) = .25880
c Excel with n = 25 and p = 18/38: P(X ≥ 15) = 1 – P(X ≤ 14) = 1 – .85645 = .14355
d Excel with n = 25 and p = 18/38: P(X ≤ 10) = .29680
7.106 a Excel with n = 100 and p = .52: P(X ≥ 50) = 1 – P(X ≤ 49) = 1 – .30815 = .69185
b Excel with n = 100 and p = .36: P(X ≤ 30) = .12519
c Excel with n = 100 and p = .06: P(X ≤ 5) = .44069
7.107 Excel with n = 20 and p = .38: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .81032 = .18968
7.108a. Excel with n = 10 and p = .23: P(X ≥ 5) = 1 – P(X ≤ 4) = 1 – .94308 = .05692

b. Excel with n = 25 and p = .23: P(X ≤ 5) = .47015
7.109 Excel with n = 50 and p = .45: P(X ≥ 19) = 1 – P(X ≤ 18) = 1 – .12735 = .87265
e −µ µ x e −2 2 0
7.110 a P(X = 0) = = = .1353
x! 0!
e −µ µ x e −2 2 3
b P(X = 3) = = = .1804
x! 3!
e −µ µ x e −2 2 5
c P(X = 5) = = = .0361
x! 5!
e −µ µ x e −.5 .5 0
7.111a P(X = 0) = = = .6065
x! 0!
e −µ µ x e −.5 .51
b P(X = 1) = = = .3033
x! 1!
209
e −µ µ x e −.5 .5 2
c P(X = 2) = = = .0758
x! 2!
7.112 a Table 2 with µ = 3.5: P(X = 0) = P(X ≤ 0) = .0302

b Table 2 with µ = 3.5: P(X ≥ 5) = 1 – P(X ≤ 4) = 1 – .7254 = .2746
c Table 2 with µ = 3.5/7: P(X = 1) = P(X ≤ 1) – P(X ≤ 0) = .9098 – .6065 = .3033
e −µ µ x e −14 / 3 (14 / 3) 5
7.113 a P(X = 5 with µ = 14/3) = = = .1734
x! 5!
e −µ µ x e −1 / 3 (1 / 3)1
b. P(X = 1 with µ = 14/3) = = = .2388
x! 1!
e −µ µ x e −2 (2) 0
7.114 a P(X = 0 with µ = 2) = = = .1353
x! 0!
e −µ µ x e −14 (14)10
b P(X = 10 with µ = 14) = = = .0663
x! 10!
7.115 a Table 2 with µ = 5: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .9682 = .0318

b Table 2 with µ = 10: P(X ≥ 20) = 1 – P(X ≤ 19) = 1 – .9965 = .0035
7.116 a Excel with µ = 30: P(X ≥ 35) = 1 – P(X ≤ 34) = 1 – .79731 = .20269
b Excel with µ = 15:P(X ≤ 12 = .26761
7.117 a Excel with µ = 1.8: P(X ≥ 3) = 1 – P(X ≤ 2) = 1 – .73062 = .26938

b Table 2 with µ = 9: P(10 ≤ X ≤ 15) = P(X ≤ 15) – P(X ≤ 9) = .9780 – .5874 = .3906
e −µ µ x e −.4 (.4) 0
7.118 P(X = 0 with µ = 80/200) = = =.6703
x! 0!
7.119 a Table 2 with µ = 5: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 – .9682= .0318

b Excel with µ = 25: P(X ≥ 25) = 1 – P(X ≤ 24) = 1 – .47340 = .52660
7.120 a Table 2 with µ = 1.5: P(X ≥ 2) = 1 – P(X ≤ 1) = 1 – .5578 = .4422

b Table 2 µ = 6: P(X < 4) = P(X ≤ 3) = .1512
210
e −µ µ x e −5 (5)1
7.121 a P(X = 1 with µ = 5) = = = .0337
x! 1!
b Table 2 with µ = 15: P(X > 20) = P(X ≥ 21) = 1 – P(X ≤ 20) = 1 – .9170 = .0830
e −µ µ x e −1.5 (1.5) 0
7.122 a P(X = 0 with µ = 1.5) = = = .2231
x! 0!
b Table 2 with µ = 4.5: P(X ≤ 5) = .7029
c Table 2 with µ = 3.0: P(X ≥ 3) = 1 – P(X ≤ 2 = 1 – .4232 = .5768
5!
7.123 P(X = 5) = (.774) 5 (1 − .774) 5−5 = .2778
5! (5 − 5)!
7.124 a E(X) = np = 40(.02) = .8

40!
b P(X = 0) = (.02)0 (1 − .02)40−0 = .4457
0!(40 − 0)!
7.125 a µ = E(X) = ∑ xP(x) = 0(.48) + 1(.35) + 2(.08) + 3(.05) + 4(.04) = .82

2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–.82) (.48) + (1–.82) (.35) + (2–.82) (.08)
2 2
+ (3–.82) (.05) + (4–.82) (.04) = 1.0876
σ= σ 2 = 1.0876 = 1.0429
e −µ µ x e −8 (8)10
7.126 a P(X = 10 with µ = 8) = = = .0993
x! 10!
b Table 2 with µ = 8: P(X > 5) = P(X ≥ 6) = 1 – P(X ≤ 5) = 1 – .1912 = .8088
c Table 2 with µ = 8: P(X < 12) = P(X ≤ 11) = .8881
7.127 a E(X) = np = 100(.15) = 15
b σ= np(1 − p) = 100(.15)(1 − .15) = 3.57
c Excel with n = 100 and p = .15: P(X ≥ 20) = 1 – P(X ≤ 19) = 1 – .89346 = .10654
7.128 Table 1 with n = 10 and p = .3: P(X > 5) = P(X ≥ 6) = 1 – P(X ≤ 5) = 1 – .9527 = .0473
7.129 a µ = E(X) = ∑ xP(x) = 0(.05) + 1(.16) + 2(.41) + 3(.27) + 4(.07) + 5(.04) = 2.27
211
2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–2.27) (.05) + (1–2.27) (.16) + (2–2.27) (.41)
2 2 2
+ (3–2.27) (.27) + (4–2.27) (.07) + (5–2.27) (.04) = 1.1971
σ= σ 2 = 1.1971 = 1.0941
7.130 Table 1 with n = 10 and p = .20: P(X ≥ 6) = 1 – P(X ≤ 5) = 1 – .9936 = .0064
10!
7.131 a P(X = 2) = (.05) 2 (1 − .05)10− 2 = .0746
2!(10 − 2)!
b Excel with n = 400 and p = .05: P(X = 25) = .04455
c .05
7.132 a Excel with n = 80 and p = .70: P(X > 65) = P(X ≥ 66) = 1 – P(X ≤ 65) = 1 – .99207 =
.00793
b E(X) = np = 80(.70) = 56
c σ= np(1 − p) = 80(.70)(1 − .70) = 4.10
7.133 a Excel with µ = 9.6: P(X ≥ 10) = 1 – P(X ≤ 9) = 1 − .5089 = .4911
b. Excel with µ = 6: P(X ≤ 5) = .4457

c Excel with µ = 2.3: P(X ≥ 3) = 1 – P(X ≤ 2) = 1 − .5960 = .4040
7.134 Table 1 with n = 25 and p = .40:

a P(X = 10) = P(X ≤ 10) – P(X ≤ 9) = .5858 – .4246 = .1612
b P(X < 5) = P(X ≤ 4) = .0095
c P(X > 15) = P(X ≥ 16) = 1 – P(X ≤ 15) = 1 – .9868 = .0132
7.135 Excel with n = 100 and p = .45:

a P(X > 50) = P(X ≥ 49) = 1 – P(X ≤ 50) = 1 – .86542 = .13458
b P(X < 44) = P(X ≤ 43) = .38277
c P(X = 45) = .07999
7.136 a. µ = E(X) = ∑ xP(x) = 0(.36) + 1(.22) + 2(.20) + 3(.09) + 4(.08) + 5(.05) = 1.46
2 2 2
σ 2 = V(X) = ∑ ( x − µ) 2
P( x ) = (0–1.46) (.36) + (1–1.46) (.22) + (2–1.46) (.20)
2 2 2
+ (3–1.46) (.09) + (4–1.46) (.08) + (5–1.46) (.05) = 2.23
σ= σ 2 = 2.23 = 1.49
212
b. µ = E(X) = ∑ xP(x) = 0(.15) + 1(.18) + 2(.23) + 3(.26) + 4(.10) + 5(.08) = 2.22
2 2 2
σ2 2
= V(X) = ∑ ( x − µ) P( x ) = (0–2.22) (.15) + (1–2.22) (.18) + (2–2.22) (.23)
2 2 2
+ (3–2.22) (.26) + (4–2.22) (.10) + (5–2.22) (.08) = 2.11
σ= σ 2 = 2.11 = 1.45
7.137 Excel with n = 25 and p = 1/3: P(X ≥ 10) = 1 – P(X ≤ 9) = 1– .69560 = .30440
7.138 p = .08755 because P(X ≥ 1) = 1– P(X = 0 with n = 10 and p = .08755) = 1– .40 = .60
7.139 Excel with n = 100 and p = .60: P(X > 50) = P(X ≥ 51) = 1 – P(X ≤ 50) = 1– .02710 =
.97290
7.140 Binomial with n = 5 and p = .01. (using Excel)

x p(x)
0 .95099
1 .04803
2 .00097
3 .00001
4 0
5 0
Case 7.1
Expected number of runs without bunting = .85.
If batter bunts:
Bases Expected Number
Outcome Probability Occupied Outs of Runs
1 .75 2nd 1 .69 .5175

2 .10 1st 1 .52 .0520
3 .10 none 2 .10 .0100
4 .05 1st and 2nd 0 1.46 .0730
Expected number of runs = .6255
Decision: Don’t bunt.
213
214
Chapter 8
(60 − 45) × 2 (75 − 60) × 2

8.1a. P(X > 45) ≈ + = .0800
50 ×15 50 ×15
(15 − 10) ×17 (30 − 15) × 7 (40 − 30) × 6
b. P(10 < X < 40) ≈ + + = .3333
50 ×15 50 ×15 50 ×15
(−15 − [−30]) × 6 (0 − [−15]) ×10 (15 − 0) ×17 (25 − 15) × 7
c. P(X < 25) ≈ + + + = .7533
50 ×15 50 ×15 50 ×15 50 ×15
(45 − 35) × 6 (60 − 45) × 2 (65 − 60) × 2
d. P(35 < X < 65) ≈ + + = .1333
50 ×15 50 ×15 50 ×15
(60 − 45) × 3 (75 − 60) × 3

8.2a. P(X > 45) ≈ + = .1200
50 ×15 50 ×15
(15 − 10) ×16 (30 − 15) × 8 (40 − 30) × 8
b. P(10 < X < 40) ≈ + + = .4800
50 ×15 50 ×15 50 ×15
(−30 − [−45]) × 5 (−15 − [−30]) × 5 (0 − [−15]) × 2 (15 − 0) ×16 (25 − 15) × 8
c. P(X < 25) ≈ + + + +
50 ×15 50 ×15 50 ×15 50 ×15 50 ×15
= .6667
(45 − 35) × 8 (60 − 45) × 3 (65 − 60) × 3
d. P(35 < X < 65) ≈ + + = .1867
50 ×15 50 ×15 50 ×15
(60 − 55) × 16 (70 − 60) × 5 (80 − 70) × 24

8.3a. P(55 < X < 80) ≈ + + = .6167
60 × 10 60 × 10 60 × 10
(70 − 65) × 5 (80 − 70) × 24 (90 − 80) × 7 (100 − 90) × 1
b. P(X > 65) ≈ + + + = .5750
60 × 10 60 × 10 60 × 10 60 × 10
(50 − 40) × 7 (60 − 50) × 16 (70 − 60) × 5 (80 − 70) × 24 (85 − 80) × 7
c. P(X < 85) ≈ + + + +
60 × 10 60 × 10 60 × 10 60 × 10 60 × 10
= .9250
(80 − 75) × 24 (85 − 80) × 7
d. P(75 < X < 85) ≈ + = .2583
60 ×10 60 ×10
215
8.4 a.
b. P(X > 25) = 0

1
c. P(10 < X < 15) = (15 − 10) = .25
20
1
d. P(5.0 < X < 5.1) = (5.1 − 5) = .005
20
1 1
8.5a. f(x) = = 20 < x < 60
(60 − 20) 40
1
b. P(35 < X < 45) = (45–35) = .25
40
216
c.
1 1
8.6 f(x) = = 30 < x < 60
(60 − 30) 30
1
a. P(X > 55) = (60 − 55) = .1667
30
1
b. P(30 < X < 40) = (40 − 30) = .3333
30
c. P(X = 37.23) = 0
1
8.7 × (60 − 30) = 7.5 ; The first quartile = 30 + 7.5 = 37.5 minutes
4
8.8 .10 × (60 − 30) = 3; The top decile = 60–3 = 57 minutes
1 1
8.9 f(x) = = 110 < x < 175
(175 − 110) 65
1
a. P(X > 150) = (175 − 150) = .3846
65
1
b. P(120 < X < 160) = (160 − 120) = .6154
65
8.10 .20(175–110) = 13. Bottom 20% lie below (110 + 13) = 123
For Exercises 8.11 to 8.14 we calculate probabilities by determining the area in a triangle. That
is,
Area in a triangle = (.5)(height)(base)
217
8.11a.
b. P(0 < X < 2) = (.5)(2–0)(1) = 1.0

c. P(X > 1) = (.5)(2 – 1)(.5) = .25
d. P(X < .5) = 1 – P(X > .5) = 1 – (.5)(.75)(2–.5) = 1 – .5625 = .4375
e. P(X = 1.5) = 0
8.12 a
b. P(2 < X < 4) = P(X < 4) – P(X < 2) = (.5)(3/8)(4–1) – (.5)(1/8)(2–1) = .5625 – .0625 = .5
c. P(X < 3) = (.5)(2/8)(3–1) = .25
218
8.13a.
1 3 1 1
b. P(1 < X < 3) = P(X < 3) – P(X < 1) = × × (3 − 0) − × × (1 − 0) = .18 – .02 = .16
2 25 2 25
c. P(4 < X < 8) = P(4 < X < 5) + P(5 < X < 8)
1 5 1 4
P(4 < X < 5)= P(X < 5) – P(X <4) = × × (5 − 0) − × × (4 − 0) = .5 – .32 = .18
2 25 2 25
1 5 1 2
P(5 < X < 8) = P(X > 5) – P(X > 8) = × × (10 − 5) − × × (10 − 8) = .5 – .08 = .42
2 25 2 25
P(4 < X < 8) = .18 + .42 = .60
d. P(X < 7) = 1 – P(X > 7)
1 3
P(X > 7) = × × (10 − 7) = .18
2 25
P(X < 7) = 1 – .18 = .82
e. P(X > 3) = 1 – P(X < 3)
1 3
P(X < 3) = × × (3 − 0) = .18
2 25
P(X > 3) = 1 – .18 = .82
8.14 a. f(x) = .10 – .005x 0 ≤ x ≤ 20

b. P(X > 10) = (.5)(.05)(20–10) = .25
c. P(6 < X < 12) = P(X > 6) – PX > 12) = (.5)(.07)(20–6) – (.5)(.04)(20–12) = .49 – .16 = .33
8.15 P( Z < 1.50) = .9332
8.16 P(Z < 1.51) = .9345
8.17 P(Z < 1.55) = .9394
8.18 P(Z < −1.59) = .0559
219
8.19 P(Z < −1.60) = .0548
8.20 P(Z < − 2.30) = .0107
8.21 P(–1.40 < Z < 0.60) = P( Z < 0.60) − P(Z < −1.40) = .7257− .0808 = .6449
8.22 P(Z > –1.44) = 1 – P(Z < −1.44) = 1 − .0749 = .9251
8.23 P(Z < 2.03) = .9788
8.24 P(Z > 1.67) = 1 – P(Z < 1.67) = 1 – .9525 = .0475
8.25 P(Z < 2.84) = ..9977
8.26 P(1.14 < Z < 2.43) = P(Z < 2.43) – P(Z < 1.14) = .9925 – .8729 = .1196
8.27 P(–0.91 < Z < –0.33) = P(Z < −.33) – P(Z < −.91) = .3707 – .1814 = .1893
8.28 P(Z > 3.09) = 1 – P(Z < 3.09) = 1 – .9990 = .0010
8.29 P(Z > 0) = 1 – P(Z < 0) = 1 − .5 = .5
8.30 P(Z > 4.0) = 0
8.31 P(Z < z .02 ) = 1 – .02 = .9800; z .02 = 2.05
8.32 P(Z < z .045 ) = 1 – .045 = .9550; z .045 = 1.70
8.33 P(Z < z .20 ) = 1 – .20 = .8000; z .20 = .84
⎛ X − µ 145 − 100 ⎞
8.34 P(X > 145) = P⎜ > ⎟ = P(Z > 2.25) = 1 – P(Z < 2.25) = 1 – .9878 = .0122
⎝ σ 20 ⎠
x−µ x − 250
8.35 P(Z < z.15 ) = 1 – .15 = .8500; z .15 = 1.04; z .15 = ; 1.04 = ; x = 291.6
σ 40
220
⎛ 800 − 1,000 X − µ 1,100 − 1,000 ⎞
8.36 P(800 < X < 1100) = P⎜ < < ⎟ = P(–.8 < Z < .4)
⎝ 250 σ 250 ⎠
= P(Z < .4) − P(Z < −.8) = .6554 − .2119 = .4435
x −µ x − 50
8.37 P(Z < −z .08 ) = .0800; −z .08 = −1.41; − z .08 = ; − 1.41 = ; x = 38.72
σ 8
⎛ 5 − 6.3 X − µ 10 − 6.3 ⎞
8.38 a P(5 < X < 10) = P⎜ < < ⎟ = P(–.59 < Z > 1.68)
⎝ 2.2 σ 2.2 ⎠
= P(Z < 1.68) − P(Z < −.59) = .9535 − .2776 = .6759
⎛ X − µ 7 − 6.3 ⎞
b P(X > 7) = P⎜ > ⎟ = P(Z > .32) = 1 – P(Z < .32) = 1 – .6255 = .3745
⎝ σ 2.2 ⎠
⎛ X − µ 4 − 6.3 ⎞
c P(X < 4) = P⎜ < ⎟ = P(Z < –1.05) = .1469
⎝ σ 2.2 ⎠
x −µ x − 6.3
8.39 P(Z < z .10 ) = 1 – .10 = .9000; z .10 = 1.28; z .10 = ; 1.28 = ; x = 9.116
σ 2.2
Calls last at least 9.116 minutes.
⎛ X − µ 5,000 − 5,100 ⎞
8.40 P(X > 5,000) = P⎜ > ⎟ = P(Z > –.5) = 1 −P(Z < −.5) = 1 − .3085 = .6915
⎝ σ 200 ⎠
x −µ x − 5100
8.41 P(Z < −z .02 ) = .02; −z .02 = −2.05; − z .02 = ; − 2.05 = ; x = 4690;
σ 200
⎛ X − µ 12,000 − 10,000 ⎞
8.42 a P(X > 12,000) = P⎜⎜ > ⎟⎟ = P(Z > .83) = 1 – P(Z < .83) = 1 – .7967 =
⎝ σ 2,400 ⎠
.2033
⎛ X − µ 9,000 − 10,000 ⎞
b P(X < 9,000) = P⎜⎜ < ⎟⎟ = P(Z < –.42) = .3372
⎝ σ 2,400 ⎠
x−µ x − 10,000
8.43 P(Z < z .001 ) = .9990; z .001 = 3.08; z .001 = ; 3.08 = ; x = 17,392
σ 2,400
⎛ X − µ 70 − 65 ⎞
8.44 a P(X > 70) = P⎜ > ⎟ = P(Z > 1.25) = 1 – P(Z < 1.25) = 1 – .8944 = .1056
⎝ σ 4 ⎠
221
⎛ X − µ 60 − 65 ⎞
b P(X < 60) = P⎜ < ⎟ = P(Z < –1.25) = .1056
⎝ σ 4 ⎠
⎛ 55 − 65 X − µ 70 − 65 ⎞
c P(55 < X < 70) = P⎜ < < ⎟ = P(–2.50 < Z < 1.25)
⎝ 4 σ 4 ⎠
= P(Z < 1.25) − P(Z < −2.50) = .8944− .0062 = .8882
⎛ X − µ 70,000 − 82,000 ⎞
8.45 a P(X < 70,000) = P⎜⎜ < ⎟⎟ = P(Z < –1.88) = .0301
⎝ σ 6,400 ⎠
⎛ X − µ 100,000 − 82,000 ⎞
b P(X > 100,000) = P⎜⎜ > ⎟⎟ = P(Z > 2.81) = 1 – P(Z < 2.81) = 1 – .9975 =
⎝ σ 6,400 ⎠
.0025
x −µ x − 32
8.46 Top 5%: P(Z < z .05 ) = 1 – .05 = .9500; z .05 = 1.645; z .05 = ; 1.645 = ;x=
σ 1.5
34.4675
x −µ x − 32
Bottom 5%: P(Z < −z .05 ) = .0500; −z .05 = −1.645; − z .05 = ; − 1.645 = ;
σ 1.5
x = 29.5325
⎛ X − µ 36 − 32 ⎞
8.47 a P(X > 36) = P⎜ > ⎟ = P(Z > 2.67) = 1 – P(Z < 2.67) = 1 – .9962 = .0038
⎝ σ 1.5 ⎠
⎛ X − µ 34 − 32 ⎞
b P(X < 34) = P⎜ < ⎟ = P(Z < 1.33) = .9082
⎝ σ 1.5 ⎠
⎛ 30 − 32 X − µ 33 − 32 ⎞
c P(30 < X < 33) = P⎜ < < ⎟ = P(–1.33 < Z < .67)
⎝ 1.5 σ 1.5 ⎠
= P(Z < .67) − P(Z < −1.33) = .7486 − .0918 = .6568
⎛ X − µ 8 − 7.2 ⎞
8.48 P(X > 8) = P⎜ > ⎟ = P (Z > 1.20) = 1 – P(Z < 1.20) = 1 – .8849 = .1151
⎝ σ .667 ⎠
x −µ x − 7.2
8.49 P(Z < z .25 ) = .7500; z .25 = .67; z .25 = ; .67 = ; x = 7.65 hours
σ .667
⎛ X − µ 10 − 7.5 ⎞
8.50 a P(X > 10) = P⎜ > ⎟ = P Z > 1.19) = 1 – P(Z < 1.19) = 1 – .8830 = .1170
⎝ σ 2.1 ⎠
222
⎛ 7 − 7.5 X − µ 9 − 7.5 ⎞
b P(7 < X < 9) = P⎜ < < ⎟ = P(–.24 < Z < .71)
⎝ 2.1 σ 2.1 ⎠
= P(Z < .71) − P(0 < Z < −.24) = .7611 − .4052 = .3559
⎛ X − µ 3 − 7.5 ⎞
c P(X < 3) = P⎜ < ⎟ = P Z < –2.14) = .0162
⎝ σ 2.1 ⎠
x −µ x − 7.5
d P(Z < – z .05 ) = .0500; −z .05 = –1.645; − z .05 = ; − 1.645 = ; x = 4.05 hours
σ 2.1
⎛ X − µ 12,000 − 11,500 ⎞
8.51 a P(X > 12,000) = P⎜ > ⎟ = P Z > .63) = 1– P(Z < .63) = 1 – .7357 =
⎝ σ 800 ⎠
.2643
⎛ X − µ 10,000 − 11,500 ⎞
b P(X < 10,000) = P⎜ < ⎟ = P(Z < –1.88) = .0301
⎝ σ 800 ⎠
x −µ x − 11,500
8.52 P(Z < – z .01 ) = .0100; −z .05 = –2.33; − z .01 = ; − 2.33 = ; x = 9,636
σ 800
⎛ 24 − 26 X − µ 28 − 26 ⎞
8.53 a P(24 < X < 28) = P⎜ < < ⎟ = P(–.80 < Z < .80)
⎝ 2.5 σ 2.5 ⎠
= P(Z < .80) – P(Z < −.80) = .7881 − .2119 = .5762
⎛ X − µ 28 − 26 ⎞
b P(X > 28) = P⎜ > ⎟ = P(Z > .80) = 1 – P(Z < .80) = 1 – .7881 = .2119
⎝ σ 2.5 ⎠
⎛ X − µ 24 − 26 ⎞
c P(X < 24) = P⎜ < ⎟ = P(Z < –.80) =.2119
⎝ σ 2.5 ⎠
⎛ X − µ 30 − 27 ⎞
8.54 a P(X > 30) = P⎜ > ⎟ = P(Z > .43) = 1 – P(Z < .43) = 1 – .6664 = .3336
⎝ σ 7 ⎠
⎛ X − µ 40 − 27 ⎞
b P(X > 40) = P⎜ > ⎟ = P(Z > 1.86) = 1 – P(Z < 1.86) = 1 – .9686 = .0314
⎝ σ 7 ⎠
⎛ X − µ 15 − 27 ⎞
c P(X < 15) = P⎜ < ⎟ = P(Z < –1.71) = .0436
⎝ σ 7 ⎠
x −µ x − 27
d P(Z < z .20 ) = 1 – .20 = .8000; z .20 = .84; z .20 = ; .84 = ; x = 32.88
σ 7
⎛ X − µ 4 − 7.5 ⎞
8.55 a P(X < 4) = P⎜ < ⎟ = P(Z < –2.92) = .0018
⎝ σ 1.2 ⎠
223
⎛ 7 − 7.5 X − µ 10 − 7.5 ⎞
b P(7 < X < 10) = P⎜ < < ⎟ = P(–.42 < Z < 2.08)
⎝ 1.2 σ 1.2 ⎠
= P(Z < 2.08) − P(Z < −.42) = .9812 − .3372 = .6440
⎛ X − µ 10 − 16.40 ⎞
8.56 a P(X < 10) = P⎜ < ⎟ = P(Z < –2.33) = .0099
⎝ σ 2.75 ⎠
x −µ x − 16.40
b P(Z < – z .10 ) = .1000; – z .10 = –1.28; − z .10 = ; − 1.28 = ; x = 12.88
σ 2.75
x −µ x − 70
8.57 A: P(Z < z .10 ) = 1– .10 = .9000; z .10 = 1.28; z .10 = ; 1.28 = ; x = 82.8
σ 10
x −µ x − 70
B: P(Z < z .40 ) = 1 –.40 = .6000; z .40 = .25; z .40 = ; .25 = ; x = 72.5
σ 10
x −µ x − 70
C: P(Z < – z .20 ) = .2000; −z .20 = −.84; − z .20 = ; − .84 = ; x = 61.6;
σ 10
x −µ x − 70
D: P(Z < – z .05 ) = .0500; −z .05 = −1.645; − z .05 = ; − 1.645 = ; x = 53.55
σ 10
x −µ x − 100
8.58 P(Z < z .02 ) = 1 – .02 = .9800; z .02 = 2.05; z .02 = ; 2.05 = ; x = 132.80
σ 16
(rounded to 133)
⎛ X − µ 70,000 − 61,823 ⎞
8.59 P(X > 70,000) = P⎜⎜ > ⎟⎟ = P(Z > .47) = 1 – P(Z < .47) = 1 – .6808
⎝ σ 17,301 ⎠
= .3192
⎛ X − µ 45,000 − 41,825 ⎞
8.60 P(X < 45,000) = P⎜⎜ < ⎟⎟ = P(Z < .24) = .5948
⎝ σ 13,444 ⎠
x −µ x − 75
8.61 P(Z < −z .01 ) = .0100; −z .01 = −2.33; − z .01 = ; − 2.33 = ; x = 56.36
σ 8
⎛ X − µ 150,000 − 99,700 ⎞
8.62 P(x > 150,000) = P⎜⎜ < ⎟⎟ = P(Z > 1.68) = 1 – P(Z < 1.68) = 1 – .9535
⎝ σ 30,000 ⎠
= .0465
224
ROP − µ ROP − 200
8.63 P(Z < z .06 ) = 1 – .06 = .9400; z .06 = 1.55; z .06 = ; 1.55 = ;
σ 30
ROP = 246.5 (rounded to 247)
x −µ x − 150
8.64 P(Z < z.20 ) = 1 – .20 = .8000; z .20 = .84; z .20 = ; .84 = ; x = 171
σ 25
x −µ x − 850
8.65 P(Z < z .30 ) = 1 – .30 = .7000; z .30 = .52; z .30 = ; .52 = ; x = 896.8
σ 90
(rounded to 897)
x −µ x − 850
8.66 P( Z < z .40 ) = 1– .40 = .6000; z .40 = .25; z .40 = ; .25 = ; .x = 872.5
σ 90
(rounded to 873)
8.67 From Exercise 7.57: µ = 65, σ 2 = 21, and σ = 4.58
⎛ X − µ 60 − 65 ⎞
P(X > 60) = P⎜ > ⎟ = P(Z > –1.09) = 1 − P(Z < −1.09) = 1 −.1379 = .8621
⎝ σ 4.58 ⎠
⎛ X − µ 150 − 145 ⎞
8.68 P(X < 150) = P⎜ < ⎟ = P(Z < .90) = .8159
⎝ σ 5.57 ⎠
⎛ X − µ 25 − 14 ⎞
8.69 a. P(X > 25) = P⎜ > ⎟ = P(Z > .61) = 1 – P(Z < .61) = 1 – .7291 = .2709
⎝ σ 18 ⎠
⎛ X − µ 0 − 14 ⎞
b. P(X < 0) = P⎜ < ⎟ = P(Z < –.78) = .2177
⎝ σ 18 ⎠
⎛ X − µ 0 − 10.60 ⎞
8.70 a. P(X < 0) = P⎜ < ⎟ = P(Z < –.73) = .2327
⎝ σ 14.56 ⎠
⎛ X − µ 20 − 10.60 ⎞
b. P(X > 20) = P⎜ > ⎟ = P(Z > .65) = 1 − P(Z < .65) = 1 − .7422 = .2578
⎝ σ 14.56 ⎠
225
8.71
8.72
8.73 a P(X > 1) = e −..5(1) = e −..5 = .6065
a P(X > .4) = e −..5(.4) = e −..2 = .8187
c P(X < .5) = 1 − e −..5(..5) = 1 − e −..25 = 1 – .7788 = .2212
d P(X < 2) = 1 − e −.5(2) = 1 − e −1 = 1 – .3679 = .6321
8.74 a P(X > 2) = e −..3(2) = e −.6 = .5488
b P(X < 4) = 1 − e −..3(4) = 1 − e −1.2 = 1 – .3012 = .6988
226
c P(1 < X < 2) = e −..3(1) − e −..3(2) = e −..3 − e −.6 = .7408 – .5488 = .1920
d P(X = 3) = 0
8.75 λ = 6 kilograms/hour = .1 kilogram/minute
P(X > 15) = e −.1(15) = e −1.5 = .2231
8.76 µ = 1 / λ = 25 hours; λ = .04 breakdowns/hour
P(X > 50) = e −.04(50) = e −2 = .1353
8.77 λ = 10 trucks/hour = .167 truck/minute
P(X > 15) = e −.167(15) = e −2.5 = .0821
8.78 µ = 1 / λ = 5 minutes; λ = .2 customer/minute
P(X < 10) = 1 − e −.2(10) = 1 − e −2 = 1– .1353 = .8647
8.79 µ = 1 / λ = 2.7 minutes; λ = .37 service/minute
P(X < 3) = 1 − e −.37(3) = 1 − e −1.11 = 1– .3296 = .6704
8.80 µ = 1 / λ = 7.5 minutes; λ = .133 service/minute
P(X < 5) = 1 − e −.133(5) = 1 − e −.665= 1– .5143 = .4857
8.81 µ = 1 / λ = 125 seconds; λ = .008 transactions/second = .48 transactions/minute
P(X > 3) = e −.48(3) = e −1.44 = .2369
8.82 µ = 1 / λ = 6 minutes; λ = .167 customers/minute
P(X > 10) = e −.167(10) = e −1.67 = .1889
8.83 a 1.341 b 1.319 c 1.988 d 1.653
8.84 a 2.750 b 1.282 c 2.132 d 2.528
8.85 a 1.3406 b 1.3195 c 1.9890 d 1.6527
8.86 a 1.6556 b 2.6810 c 1.9600 d 1.6602
227
8.87 a .0189 b .0341 c .0927 d .0324
8.88 a .1744 b .0231 c .0251 d .0267
8.89 a 9.24 b 136 c 9.39 d 37.5
8.90 a 17.3 b 50.9 c 2.71 d 53.5
8.91 a 73.3441 b 102.946 c 16.3382 d 24.7690
8.92 a 33.5705 b 866.911 c 24.3976 d 261.058
8.93 a .2688 b 1.0 c .9903 d 1.0
8.94 a .4881 b .9158 c .9988 d .9077
8.95 a 4.35 b 8.89 c 3.29 d 2.50
8.96 a 2.84 b 1.93 c 3.60 d 3.37
8.97 a 1.4857 b 1.7633 c 1.8200 d 1.1587
8.98 a 1.5204 b 1.5943 c 2.8397 d 1.1670
8.99 a .0510 b .1634 c .0222 d .2133
8.100 a .1050 b .1576 c .0001 d .0044
228
Chapter 9
9.1a. 1/6
b. 1/6
9.2 a P( X = 1) =P(1,1)= 1/36
b P( X = 6) = P(6,6) = 1/36
9.3a P( X = 1) = (1/6) 5 = .0001286
b P( X = 6) = (1/6) 5 = .0001286
9.4 The variance of X is smaller than the variance of X.
9.5 The sampling distribution of the mean is normal with a mean of 40 and a standard deviation of
12/ 100 = 1.2.
9.6 No, because the sample mean is approximately normally distributed.
⎛ X − µ 1050 − 1000 ⎞
9.7 a P( X > 1050) = P⎜ > ⎟ = P(Z > 1.00) = 1 – P(Z < 1.00) = 1 – .8413
⎜ ⎟
⎝ σ / n 200 / 16 ⎠
= .1587
⎛ X − µ 960 − 1000 ⎞
b P( X < 960) = P⎜ < ⎟ = P(Z < –.80) = .2119
⎜ ⎟
⎝ σ / n 200 / 16 ⎠
⎛ X − µ 1100 − 1000 ⎞
c P( X > 1100) = P⎜ > ⎟ = P(Z > 2.00) = 1 – P(Z < 2.00) = 1 – .9772 = .0228
⎜ ⎟
⎝ σ / n 200 / 16 ⎠
⎛ X − µ 1050 − 1000 ⎞
9.8 a P( X > 1050) = P⎜ > ⎟ = P(Z > 1.25) = 1 – P(Z < 1.25) = 1 – .8944
⎜ ⎟
⎝ σ / n 200 / 25 ⎠
= .1056
⎛ X − µ 960 − 1000 ⎞
b P( X < 960) = P⎜ < ⎟ = P(Z < –1.00) = .1587
⎜ ⎟
⎝ σ / n 200 / 25 ⎠
⎛ X − µ 1100 − 1000 ⎞
c P( X > 1100) = P⎜ > ⎟ = P(Z > 2.50) = 1 – P(Z < 2.50) = 1 – .9938 = .0062
⎜ ⎟
⎝ σ / n 200 / 25 ⎠
229
⎛ X − µ 1050 − 1000 ⎞
9.9 a P( X > 1050) = P⎜ > ⎟ = P(Z > 2.50) = 1 – P(Z < 2.50) = 1 – .9938
⎜ ⎟
⎝ σ / n 200 / 100 ⎠
= .0062
⎛ X − µ 960 − 1000 ⎞
b P( X < 960) = P⎜ < ⎟ = P(Z < –2.00) = .0228
⎜ ⎟
⎝ σ / n 200 / 100 ⎠
⎛ X − µ 1100 − 1000 ⎞
c P( X > 1100) = P⎜ > ⎟ = P(Z > 5.00) = 0
⎜ ⎟
⎝ σ / n 200 / 100 ⎠
⎛ 49 − 50 X − µ 52 − 50 ⎞
9.10 a P(49 < X < 52) = P⎜⎜ < < ⎟ = P(–.40 < Z < .80)
⎟
⎝ 5 / 4 σ / n 5 / 4 ⎠
= P(Z < .80) − P(Z < −.40) = .7881 −.3446 = .4435
⎛ 49 − 50 X − µ 52 − 50 ⎞
b P(49 < X < 52) = P⎜ < < ⎟ = P(–.80 < Z < 1.60)
⎜ ⎟
⎝ 5 / 16 σ / n 5 / 16 ⎠
= P(Z < 1.60) − P(Z < −.80) = .9452 −.2119 = .7333
⎛ 49 − 50 X − µ 52 − 50 ⎞
c P(49 < X < 52) = P⎜ < < ⎟ = P(–1.00 < Z < 2.00)
⎜ ⎟
⎝ 5 / 25 σ / n 5 / 25 ⎠
= P(Z < 2.00) − P(Z < −1.00) = .9772 −.1587 = .8185
⎛ 49 − 50 X − µ 52 − 50 ⎞
9.11 a P(49 < X < 52) = P⎜⎜ < < ⎟ = P(–.20 < Z < .40)
⎟
⎝ 10 / 4 σ / n 10 / 4 ⎠
= P(Z < .40) − P(Z < −.20) = .6554 −.4207= .2347
⎛ 49 − 50 X −µ 52 − 50 ⎞
b P(49 < X < 52) = P⎜ < < ⎟ = P(–.40 < Z < .80)
⎜ ⎟
⎝ 10 / 16 σ / n 10 / 16 ⎠
= P(Z < .80) − P(Z < −.40) = .7881 −.3446= .4435
⎛ 49 − 50 X −µ 52 − 50 ⎞
c P(49 < X < 52) = P⎜ < < ⎟ = P(–.50 < Z < 1.00)
⎜ ⎟
⎝ 10 / 25 σ / n 10 / 25 ⎠
= P(Z < 1.00) − P(Z < −.50) = .8413 −.3085= .5328
⎛ 49 − 50 X − µ 52 − 50 ⎞
9.12 a P(49 < X < 52) = P⎜ < < ⎟ = P(–.10 < Z < .20)
⎜ ⎟
⎝ 20 / 4 σ / n 20 / 4 ⎠
= P(Z < .20) − P(Z < −.10) = .5793 −.4602= .1191
⎛ 49 − 50 X −µ 52 − 50 ⎞
b P(49 < X < 52) = P⎜ < < ⎟ = P(–.20 < Z < .40)
⎜ ⎟
⎝ 20 / 16 σ / n 20 / 16 ⎠
= P(Z < .40) − P(Z < −.20) = .6554 −.4207= .2347
230
⎛ 49 − 50 X −µ 52 − 50 ⎞
c P(49 < X < 52) = P⎜ < < ⎟ = P(–.25 < Z < .50)
⎜ ⎟
⎝ 20 / 25 σ / n 20 / 25 ⎠
= P(Z < .50) − P(Z < −.25) = .6915 −.4013= .2902
N−n 1,000 − 100

9.13 a = = .9492
N −1 1,000 − 1
N−n 3,000 − 100

b = = .9834
N −1 3,000 − 1
N−n 5,000 − 100

c = = .9900
N −1 5,000 − 1
d. The finite population correction factor is approximately 1.
σ N−n 500 10,000 − 1,000

9.14 a σ x = = = 15.00
n N −1 1,000 10,000 − 1
σ N−n 500 10,000 − 500

b σx = = = 21.80
n N −1 500 10,000 − 1
σ N−n 500 10,000 − 100

c σx = = = 49.75
n N −1 100 10,000 − 1
⎛ X − µ 66 − 64 ⎞
9.15 a P(X > 66) = P⎜ > ⎟ = P(Z > 1.00) = 1 – P(Z < 1.00) = 1 – .8413 = .1587
⎝ σ 2 ⎠
⎛ X − µ 66 − 64 ⎞
b P( X > 66) = P⎜⎜ > ⎟ = P(Z > 2.00) = 1 – P(Z < 2.00) = 1 – .9772 = .0228
⎟
⎝ σ / n 2 / 4 ⎠
⎛ X − µ 66 − 64 ⎞
c P( X > 66) = P⎜ > ⎟ = P(Z > 10.00) = 0
⎜ ⎟
⎝ σ / n 2 / 100 ⎠
9.16 We can answer part (c) and possibly part (b) depending on how nonnormal the population is.
⎛ X − µ 120 − 117 ⎞
9.17 a P(X > 120) = P⎜ > ⎟ = P(Z > 0.58) = 1 – P(Z < .58) = 1 – .7190 = .2810
⎝ σ 5.2 ⎠
⎛ X − µ 120 − 117 ⎞
b P( X > 120) = P⎜⎜ > ⎟ = P(Z > 1.15) = 1 – P(Z < 1.15) = 1 – .8749 = .1251
⎟
⎝ σ / n 5.2 / 4 ⎠
4 4
c [P(X >120)] =[.2810] = .00623
231
⎛ X − µ 60 − 52 ⎞
9.18 a P(X > 60) = P⎜ > ⎟ = P(Z > 1.33) = 1 – P(Z < 1.33) = 1 – .9082 = .0918
⎝ σ 6 ⎠
⎛ X − µ 60 − 52 ⎞
b P( X > 60) = P⎜ > ⎟ = P(Z > 2.31) = 1 – P(Z < 2.31) = 1 – .9896 = .0104
⎜ ⎟
⎝ σ / n 6 / 3 ⎠
3 3
c [P(X >60)] =[.0918] = .00077
⎛ X − µ 12 − 10 ⎞
9.19 a P(X > 12) = P⎜ > ⎟ = P(Z > .67) = 1 – P(Z < .67) = 1 – .7486 = .2514
⎝ σ 3 ⎠
⎛ X − µ 11 − 10 ⎞
b P( X > 275 / 25) = P( X > 11) = P⎜ > ⎟ = P(Z > 1.67) = 1 – P(Z < 1.67)
⎜ ⎟
⎝ σ / n 3 / 25 ⎠
= 1 – .9525 = 0475
⎛ X − µ 75 − 78 ⎞
9.20 a P(X < 75) = P⎜ < ⎟ = P(Z < –.50) = .3085
⎝ σ 6 ⎠
⎛ X − µ 75 − 78 ⎞
b P( X < 75) = P⎜ < ⎟ = P(Z < –3.54) = 1 – P(Z < 3.54) = 1 – 1 = 0
⎜ ⎟
⎝ σ / n 6 / 50 ⎠
⎛ X − µ 7 − 6 ⎞
9.21 a P(X > 7) = P⎜ > ⎟ = P(Z > .67) = 1 – P(Z < .67) = 1 – .7486 = .2514
⎝ σ 1.5 ⎠
⎛ X − µ 7 − 6 ⎞
b P( X > 7) = P⎜ > ⎟ = P(Z > 1.49) = 1 – P(Z < 1.49) = 1 – .9319 = .0681
⎜ ⎟
⎝ σ / n 1.5 / 5 ⎠
5 5
c [P(X >7)] =[.2514] = .00100
⎛ X − µ 5.97 − 6.05 ⎞
9.22 a P( X < 5.97) = P⎜ < ⎟ = P(Z < –2.67) =.0038
⎜ ⎟
⎝ σ / n .18 / 36 ⎠
b It appears to be false.
⎛ X − µ 625 − 600 ⎞
9.23 P( X > 10,000 / 16) = P( X > 625) = P⎜ > ⎟ = P(Z > .50) = 1 – P(Z < .50)
⎜ ⎟
⎝ σ / n 200 / 16 ⎠
= 1 – .6915 = .3085
9.24 The professor needs to know the mean and standard deviation of the population of the
weights of elevator users and that the distribution is not extremely nonnormal.
232
⎛ X − µ 71.25 − 75 ⎞
9.25 P( X > 1,140 / 16) = P( X > 71.25) = P⎜ > ⎟ = P(Z > –1.50)
⎜ ⎟
⎝ σ / n 10 / 16 ⎠
= 1 − P(Z < −1.50) = 1 − 0668 = .9332
⎛ X − µ 5 − 4.8 ⎞
9.26 P(Total time > 300) = P( X > 300 / 60) = P(X > 5) = P⎜ > ⎟ = P(Z > 1.19)
⎜ ⎟
⎝ σ / n 1.3 / 60 ⎠
= 1 – P(Z < 1.19) = 1 – .8830 = .1170
9.27 No because the central limit theorem says that the sample mean is approximately normally
distributed.
⎛ X − µ 1.92 − 2.0 ⎞
9.28 P(Total number of cups > 240) = P( X > 240 / 125) = P( X > 1.92) = P⎜ > ⎟
⎜ ⎟
⎝ σ / n .6 / 125 ⎠
= P(Z > –1.49) = 1 − P(Z < −1.49) = 1 − .0681 = .9319
⎛ X − µ 300 − 275 ⎞
9.29 P(Total number of faxes > 1500) = P( X > 1500 / 5) = P( X > 300) = P⎜ > ⎟
⎜ ⎟
⎝ σ / n 75 / 5 ⎠
= P(Z > .75) = 1 – P(Z < .75) = 1 – .7734 = .2266
⎛ P̂ − p .60 − .5 ⎞
9.30a P( P̂ > .60) = P⎜ > ⎟ = P(Z > 3.46) = 0
⎜ p(1 − p) / n (.5)(1 − .5) / 300 ⎟⎠
⎝
⎛ P̂ − p .60 − .55 ⎞
b. P( P̂ > .60) = P⎜ > ⎟ = P(Z > 1.74) = 1 – P(Z < 1.74)
⎜ p(1 − p) / n (.55)(1 − .55) / 300 ⎟⎠
⎝
= 1 – .9591 = .0409
⎛ P̂ − p .60 − .6 ⎞
c. P( P̂ > .60) = P⎜ > ⎟ = P(Z > 0) = 1 – P(Z < 0) = 1 − .5 = .5
⎜ p(1 − p) / n (.6)(1 − .6) / 300 ⎟⎠
⎝
⎛ P̂ − p .22 − .25 ⎞
9.31a P( P̂ < .22) = P⎜ > ⎟ = P(Z < –1.55) = .0606
⎜ p(1 − p) / n (.25)(1 − .25) / 500 ⎟⎠
⎝
⎛ P̂ − p .22 − .25 ⎞
b. P( P̂ < .22) = P⎜ > ⎟ = P(Z < –1.96) = .0250
⎜ p(1 − p) / n (.25)(1 − .25) / 800 ⎟⎠
⎝
⎛ P̂ − p .22 − .25 ⎞
c. P( P̂ < .22) = P⎜ > ⎟ = P(Z < –2.19) = .0143
⎜ p(1 − p) / n (.25)(1 − .25) / 1000 ⎟⎠
⎝
233
⎛ P̂ − p .75 − .80 ⎞
9.32 P( P̂ < .75) = P⎜ > ⎟ = P(Z < –1.25) = .1056
⎜ p(1 − p) / n (.80)(1 − .80) / 100 ⎟⎠
⎝
⎛ P̂ − p .35 − .40 ⎞
9.33 P( P̂ > .35)= P⎜ > ⎟ = P(Z > –.79) = .1 − P(Z < −.79)
⎜ p(1 − p) / n (.40)(1 − .40) / 60 ⎟⎠
⎝
1 − .2148= .7852
⎛ P̂ − p .49 − .55 ⎞
9.34 P( P̂ < .49) = P⎜ < ⎟ = P(Z < –2.70) = .0035
⎜ p(1 − p) / n (.55)(1 − .55) / 500 ⎟⎠
⎝
⎛ P̂ − p .04 − .02 ⎞
9.35 P( P̂ > .04)= P⎜ > ⎟ = P(Z > 4.04) = 1 – P(Z < 4.04) = 1 – 1=
⎜ p(1 − p) / n (.02)(1 − .02) / 800 ⎟⎠
⎝
0;
The defective rate appears to be larger than 2%.
⎛ P̂ − p .50 − .53 ⎞
9.36 a P( P̂ < .50) = P⎜ < ⎟ = P(Z < –1.20) =.1151; the claim may
⎜ p(1 − p) / n (.53)(1 − .53) / 400 ⎟⎠
⎝
be true
⎛ P̂ − p .50 − .53 ⎞
b P( P̂ < .50) = P⎜ < ⎟ = P(Z < –1.90) = .0287; the claim appears
⎜ p(1 − p) / n (.53)(1 − .53) / 1,000 ⎟⎠
⎝
to be false
⎛ P̂ − p .10 − .14 ⎞
9.37 P( P̂ > .10) = P⎜ > ⎟ = P(Z > –1.15) = 1 – P(Z < – 1.15)
⎜ p(1 − p) / n (.14)(1 − .14) / 100 ⎟⎠
⎝
= 1 – .1251 = .8749
⎛ P̂ − p .05 − .03 ⎞
9.38 P( P̂ > .05)= P⎜ > ⎟ = P(Z > 2.34) = 1 – P(Z < 2.34)
⎜ p(1 − p) / n (.03)(1 − .03) / 400 ⎟⎠
⎝
= 1 – .9904 = .0096; the commercial appears to be dishonest
⎛ P̂ − p .32 − .30 ⎞
9.39 P( P̂ > .32) = P⎜ > ⎟ = P(Z > 1.38) = 1 – P(Z < 1.38)
⎜ p(1 − p) / n (.30)(1 − .30) / 1,000 ⎟⎠
⎝
= 1 – .9162 = .0838
234
⎛ P̂ − p .45 − .50 ⎞
9.40 a P( P̂ < .45) = P⎜ < ⎟ = P(Z < –2.45) = .0071
⎜ p(1 − p) / n (.50)(1 − .50) / 600 ⎟⎠
⎝
b The claim appears to be false.
⎛ P̂ − p .75 − .80 ⎞
9.41 P( P̂ < .75) = P⎜ < ⎟ = P(Z < –2.34) = .0096
⎜ p(1 − p) / n (.80)(1 − .80) / 350 ⎟⎠
⎝
⎛ P̂ − p .70 − .75 ⎞
9.42 P( P̂ < .70) = P⎜ < ⎟ = P(Z < –2.48) = .0066
⎜ p(1 − p) / n (.75)(1 − .75) / 460 ⎟⎠
⎝
⎛ P̂ − p .28 − .25 ⎞
9.43 P( P̂ > .28) = P⎜ > ⎟ = P(Z > 2.40) = 1 – P(Z < 2.40)
⎜ p(1 − p) / n (.25)(1 − .25) / 1200 ⎟⎠
⎝
= 1 – .9918 = .0082
9.44 The claim appears to be false.
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 25 − (280 − 270) ⎟
9.45 P( X1 − X 2 > 25) = P⎜ 1 2 1 2
> ⎟ = P(Z > 1.21)
2 2
⎜ σ1 σ 2 25 2 30 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 10 10 ⎟⎠
⎝
= 1 – P(Z < 1.21) = 1 – .8869 = .1131
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 25 − (280 − 270) ⎟
9.46 P( X1 − X 2 > 25) = P⎜ 1 2 1 2
> ⎟ = P(Z > 2.72)
2 2
⎜ σ1 σ 2 25 2 30 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 50 50 ⎟⎠
⎝
= 1 – P(Z < 2.72) = 1 – .9967 = .0033
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 25 − (280 − 270) ⎟
9.47 P( X1 − X 2 > 25) = P⎜ 1 2 1 2
> ⎟ = P(Z > 3.84)
2 2
⎜ σ1 σ 2 25 2 30 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 100 100 ⎟⎠
⎝
= 1 – P(Z < 3.84) = 1 – 1= 0
235
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (40 − 38) ⎟
9.48 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > –1.00) = 1 – P(Z < –1.00)
2 2
⎜ σ1 σ 2 6 2 8 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 25 25 ⎟⎠
⎝
= 1 – .1587 = .8413
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (40 − 38) ⎟
9.49 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > –.50) = 1 – P(Z < –.50)
2 2
⎜ σ1 σ 2 12 2 16 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 25 25 ⎟⎠
⎝
= 1 – .3085 = .6915
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (140 − 138) ⎟
9.50 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > –1.00)
⎜ 2 2 2 2 ⎟
σ1 σ 2 6 8
⎜ + + ⎟
⎜ n1 n 2 25 25 ⎟⎠
⎝
= 1 – P(Z < –1.00) = 1 – .1587 = .8413
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (75 − 65) ⎟
9.51 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > –.77) = 1 – P(Z < −.77)
2 2
⎜ σ1 σ 2 20 2 212 ⎟
⎜ + + ⎟
⎜ n1 n 2 5 5 ⎟⎠
⎝
= 1 – .2206 = .7794
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (73 − 77) ⎟
9.52 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > .51) = 1 – P(Z < .51)
⎜ 2 2 2 2 ⎟
σ1 σ 2 12 10
⎜ + + ⎟
⎜ n n 4 4 ⎟⎠
⎝ 1 2
= 1 – .6950 = .3050
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (18 − 15) ⎟
9.53 P( X1 − X 2 > 0) = P⎜ 1 2 1 2
> ⎟ = P(Z > –2.24) = 1– P(Z < –2.24)
2 2
⎜ σ1 σ 2 3 2 3 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 10 10 ⎟⎠
⎝
= 1 – .0125 = .9875
236
⎛ ⎞
⎜ ⎟
⎜ ( X − X ) − (µ − µ ) 0 − (10 − 15) ⎟
9.54 P( X1 − X 2 < 0) = P⎜ 1 2 1 2
< ⎟ = P(Z < 5.89) = 1
2 2
⎜ σ1 σ 2 3 2 3 2 ⎟
⎜ + + ⎟
⎜ n1 n 2 25 25 ⎟⎠
⎝
237
238
Chapter 10
10.1 A point estimator is a single value; an interval estimator is a range of values.
10.2 An unbiased estimator of a parameter is an estimator whose expected value equals the
parameter.
10.3
10.4
10.5 An unbiased estimator is consistent if the difference between the estimator and the parameter
grows smaller as the sample size grows.
10.6
239
10.7 If there are two unbiased estimators of a parameter, the one whose variance is smaller is
relatively efficient.
10.8
10.9 a. x ± z α / 2 σ / n = 100 ± 1.645(25/ 50 ) = 100 ± 5.82; LCL = 94.18, UCL = 105.82
b. x ± z α / 2 σ / n = 100 ± 1.96(25/ 50 ) = 100 ± 6.93; LCL = 93.07, UCL = 106.93
c. x ± z α / 2 σ / n = 100 ± 2.575(25/ 50 ) = 100 ± 9.11; LCL = 90.89, UCL = 109.11
d. The interval widens.
10.10 a. x ± z α / 2 σ / n = 200 ± 1.96(50/ 25 ) = 200 ± 19.60; LCL = 180.40, UCL = 219.60
b. x ± z α / 2 σ / n = 200 ± 1.96(25/ 25 ) = 200 ± 9.80; LCL = 190.20, UCL = 209.80
c. x ± z α / 2 σ / n = 200 ± 1.96(10/ 25 ) = 200 ± 3.92; LCL = 196.08, UCL = 203.92
d. The interval narrows.
10.11 a. x ± z α / 2 σ / n = 80 ± 1.96(5/ 25 ) = 80 ± 1.96; LCL = 78.04, UCL = 81.96
b. x ± z α / 2 σ / n = 80 ± 1.96(5/ 100 ) = 80 ± .98; LCL = 79.02, UCL = 80.98
c. x ± z α / 2 σ / n = 80 ± 1.96(5/ 400 ) = 80 ± .49; LCL = 79.51, UCL = 80.49
10.12 a. x ± z α / 2 σ / n = 500 ± 2.33(12/ 50 ) = 500 ± 3.95; LCL = 496.05, UCL = 503.95
b. x ± z α / 2 σ / n = 500 ± 1.96(12/ 50 ) = 500 ± 3.33; LCL = 496.67, UCL = 503.33
c. x ± z α / 2 σ / n = 500 ± 1.645(12/ 50 ) = 500 ± 2.79; LCL = 497.21, UCL = 502.79
240
10.13 a. x ± z α / 2 σ / n = 500 ± 2.575(15/ 25 ) = 500 ± 7.73; LCL = 492.27, UCL = 507.73
b. x ± z α / 2 σ / n = 500 ± 2.575(30/ 25 ) = 500 ± 15.45; LCL = 484.55, UCL = 515.45
c. x ± z α / 2 σ / n = 500 ± 2.575(60/ 25 ) = 500 ± 30.91; LCL = 469.09, UCL = 530.91
10.14 a. x ± z α / 2 σ / n = 10 ± 1.645(5/ 100 ) = 10 ± .82; LCL = 9.18, UCL = 10.82
b. x ± z α / 2 σ / n = 10 ± 1.645(5/ 25 ) = 10 ± 1.64; LCL = 8.36, UCL = 11.64
c. x ± z α / 2 σ / n = 10 ± 1.645(5/ 10 ) = 10 ± 2.60; LCL = 7.40, UCL = 12.60
10.15 a. x ± z α / 2 σ / n = 100 ± 1.96(20/ 25 ) = 100 ± 7.84; LCL = 92.16, UCL = 107.84
b. x ± z α / 2 σ / n = 200 ± 1.96(20/ 25 ) = 200 ± 7.84; LCL = 192.16, UCL = 207.84
c. x ± z α / 2 σ / n = 500 ± 1.96(20/ 25 ) = 500 ± 7.84; LCL = 492.16, UCL = 507.84
d. The width of the interval is unchanged.
10.16 a. x ± z α / 2 σ / n = 400 ± 2.575(5/ 100 ) = 400 ± 1.29; LCL = 398.71, UCL = 401.29
b. x ± z α / 2 σ / n = 200 ± 2.575(5/ 100 ) = 200 ± 1.29; LCL = 198.71, UCL = 201.29
c. x ± z α / 2 σ / n = 100 ± 2.575(5/ 100 ) = 100 ± 1.29; LCL = 98.71, UCL = 101.29
d. The width of the interval is unchanged.
10.17 Yes, because the expected value of the sample median is equal to the population mean.
10.18 The variance decreases as the sample size increases, which means that the difference
between the estimator and the parameter grows smaller as the sample size grows larger.
10.19 Because the variance of the sample mean is less than the variance of the sample median, the
sample mean is relatively more efficient than the sample median.
1.2533 σ 1.2533(12)
10.20 a sample median ± z α / 2 = 500 ± 1.645 = 500 ± 3.50
n 50
241
b. The 90% confidence interval estimate of the population mean using the sample mean is 500
± 2.79.
The 90% confidence interval of the population mean using the sample median is wider than that
using the sample mean because the variance of the sample median is larger. The median is
calculated by placing all the observations in order. Thus, the median loses the potential
information contained in the actual values in the sample. This results in a wider interval estimate.
10.21 x ± z α / 2 σ / n = 6.89 ± 1.645(2/ 9 ) = 6.89 ± 1.10; LCL = 5.79, UCL = 7.99
10.22 x ± z α / 2 σ / n = 43.75 ± 1.96(10/ 8 ) = 43.75 ± 6.93; LCL = 36.82, UCL = 50.68
We estimate that the mean age of men who frequent bars lies between 36.82 and 50.68. This type
of estimate is correct 95% of the time.
10.23 x ± z α / 2 σ / n = 22.83 ± 1.96(12/ 12 ) = 22.83 ± 6.79; LCL = 16.04, UCL = 29.62
10.24 x ± z α / 2 σ / n = 9.85 ± 1.645(8/ 20 ) = 9.85 ± 2.94; LCL = 6.91, UCL = 12.79
10.25 x ± z α / 2 σ / n = 68.6 ± 1.96(15/ 15 ) = 68.6 ± 7.59; LCL = 61.01, UCL = 76.19
We estimate that the mean number of cars sold annually by all used car salespersons lies between
61.01 and 76.19. This type of estimate is correct 95% of the time.
10.26 x ± z α / 2 σ / n = 16.9 ± 2.575(5/ 10 ) = 16.9 ± 4.07; LCL = 12.83, UCL = 20.97
10.27 x ± z α / 2 σ / n = 147.33 ± 1.96(40/ 15 ) = 147.33 ± 20.24; LCL = 127.09, UCL = 167.57
10.28 x ± z α / 2 σ / n = 13.15 ± 1.645(6/ 13 ) = 13.15 ± 2.74; LCL = 10.41, UCL = 15.89
10.29 x ± z α / 2 σ / n = 75.625 ± 2.575(15/ 16 ) = 75.625 ± 9.656; LCL = 65.969, UCL =
85.281
10.30 x ± z α / 2 σ / n = 252.38 ± 1.96(30/ 400 ) = 252.38 ± 2.94; LCL = 249.44, UCL =
255.32
242
10.31 x ± z α / 2 σ / n = 1,810.16 ± 1.96(400/ 64 ) = 1,810.16 ± 98.00; LCL = 1,712.16,
UCL = 1,908.16
10.32 x ± z α / 2 σ / n = 12.10 ± 1.645(2.1/ 200 ) = 12.10 ± .24; LCL = 11.86, UCL = 12.34.
We estimate that the mean rate of return on all real estate investments lies between 11.86% and
12.34%. This type of estimate is correct 90% of the time.
10.33 x ± z α / 2 σ / n = 10.21 ± 2.575(2.2/ 100 ) = 10.21 ± .57; LCL = 9.64, UCL = 10.78
10.34 x ± z α / 2 σ / n = .510 ± 2.575(.1/ 250 ) = .510 ± .016; LCL = .494, UCL = .526. We
estimate that the mean growth rate of this type of grass lies between .494 and .526 inch . This type
of estimate is correct 99% of the time.
10.35 x ± z α / 2 σ / n = 26.81 ± 1.96(1.3/ 50 ) = 26.81 ± .36; LCL = 26.45, UCL = 27.17. We
estimate that the mean time to assemble a cell phone lies between 26.45 and 27.17 minutes. This
type of estimate is correct 95% of the time.
10.36 x ± z α / 2 σ / n = 19.28 ± 1.645(6/ 250 ) = 19.28 ± .62; LCL = 18.66, UCL = 19.90. We
estimate that the mean leisure time per week of Japanese middle managers lies between 18.66 and
19.90 hours. This type of estimate is correct 90% of the time.
10.37 x ± z α / 2 σ / n = 15.00 ± 2.575(2.3/ 100 ) = 15.00 ± .59; LCL = 14.41, UCL = 15.59.
We estimate that the mean pulse-recovery time lies between 14.41 and 15.59 minutes. This type of
estimate is correct 99% of the time.
10.38 x ± z α / 2 σ / n = 585,063 ± 1.645(30,000/ 80 ) = 585,063 ± 5,518; LCL = 579,545,
UCL = 590,581. We estimate that the mean annual income of all company presidents lies between
$579,545 and $590,581. This type of estimate is correct 90% of the time.
10.39 x ± z α / 2 σ / n = 14.98 ± 1.645(3/ 250 ) = 14.98 ± .31; LCL = 14.67, UCL = 15.29
10.40 x ± z α / 2 σ / n = 27.19 ± 1.96(8/ 100 ) = 27.19 ± 1.57; LCL = 25.62, UCL = 28.76
243
2 2
⎛ z σ ⎞ ⎛ 1.645 × 50 ⎞
10.41 a. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 68
⎝ B ⎠ ⎝ 10 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.645 × 100 ⎞
b. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 271
⎝ B ⎠ ⎝ 10 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.96 × 50 ⎞
c. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 97
⎝ B ⎠ ⎝ 10 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.645 × 50 ⎞
d. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 17
⎝ B ⎠ ⎝ 20 ⎠
10.42 a The sample size increases.

b The sample size increases.
c The sample size decreases.
2 2
⎛ z σ ⎞ ⎛ 2.575 × 250 ⎞
10.43 a. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 166
⎝ B ⎠ ⎝ 50 ⎠
2 2
⎛ z σ ⎞ ⎛ 2.575 × 50 ⎞
b. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 7
⎝ B ⎠ ⎝ 50 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.96 × 250 ⎞
c. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 97
⎝ B ⎠ ⎝ 50 ⎠
2 2
⎛ z σ ⎞ ⎛ 2.575 × 250 ⎞
d. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 4,145
⎝ B ⎠ ⎝ 10 ⎠
10.44 a The sample size decreases.

b The sample size decreases.
c The sample size increases.
2 2
⎛ z σ ⎞ ⎛ 1.645 × 10 ⎞
10.45 a. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 271
⎝ B ⎠ ⎝ 1 ⎠
b. 150 ± 1
σ 5
10.46 a. x ± z α / 2 = 150 ± 1.645 = 150 ± .5
n 271
σ 20
b. x ± z α / 2 = 150 ± 1.645 = 150 ± 2
n 271
244
10.47 a. The width of the confidence interval estimate is equal to what was specified.
b. The width of the confidence interval estimate is smaller than what was specified.
c. The width of the confidence interval estimate is larger than what was specified.
2 2
⎛ z σ ⎞ ⎛ 1.96 × 200 ⎞
10.48 a. n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 1,537
⎝ W ⎠ ⎝ 10 ⎠
b. 500 ± 10
σ 100
10.49 a. x ± z α / 2 = 500 ± 1.96 = 500 ± 5
n 1537
σ 400
b. x ± z α / 2 = 500 ± 1.96 = 500 ± 20
n 1537
10.50 a The width of the confidence interval estimate is equal to what was specified.
b The width of the confidence interval estimate is smaller than what was specified.
c The width of the confidence interval estimate is larger than what was specified.
2 2
⎛ z σ ⎞ ⎛ 1.645 × 10 ⎞
10.51 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 68
⎝ B ⎠ ⎝ 2 ⎠
2 2
⎛ z σ ⎞ ⎛ 2.575 × 360 ⎞
10.52 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 2,149
⎝ B ⎠ ⎝ 20 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.96 × 12 ⎞
10.53 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 139
⎝ B ⎠ ⎝ 2 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.645 × 20 ⎞
10.54 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 1,083
⎝ B ⎠ ⎝ 1 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.96 × 25 ⎞
10.55 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 97
⎝ B ⎠ ⎝ 5 ⎠
2 2
⎛ z σ ⎞ ⎛ 1.96 × 15 ⎞
10.56 n = ⎜⎜ α / 2 ⎟⎟ = ⎜ ⎟ = 217
⎝ B ⎠ ⎝ 2 ⎠
245
246
Chapter 11
11.1 H 0 : The drug is not safe and effective

H1 : The drug is safe and effective
11.2 H 0 : I will complete the Ph.D.

H1 : I will not be able to complete the Ph.D.
11.3 H 0 : The batter will hit one deep

H1 : The batter will not hit one deep
11.4 H 0 : Risky investment is more successful

H1 : Risky investment is not more successful
11.5 H1 : The plane is on fire
H1 : The plane is not on fire
11.6 The defendant in both cases was O. J. Simpson. The verdicts were logical because in the
criminal trial the amount of evidence to convict is greater than the amount of evidence required in
a civil trial. The two juries concluded that there was enough (preponderance of) evidence in the
civil trial, but not enough evidence (beyond a reasonable doubt) in the criminal trial.
All p-values and probabilities of Type II errors were calculated manually using Table 3 in
Appendix B.
11.7 Rejection region: z < −z.005 = −2.575or z > z.005 = 2.575

x −µ 980 − 1000
z= = = −1.00
σ/ n 200 / 100
p-value = 2P(Z < –1.00) = 2(.1587) = .3174
There is not enough evidence to infer that µ ≠ 1000.
247
11.8 Rejection region: z > z.03 = 1.88
x −µ 51 − 50
z= = = .60
σ/ n 5/ 9
p-value = P(Z > .60) = 1 – .7257 = .2743
There is not enough evidence to infer that µ > 50.
11.9 Rejection region: z < −z.10 = −1.28
x −µ 14.3 − 15
z= = = −1.75
σ/ n 2 / 25
p-value = P(Z < –1.75) = .0401
There is enough evidence to infer that µ < 15.
11.10 Rejection region: z < −z.025 = −1.96 or z > z.025 = 1.96

x −µ 100 − 100
z= = =0
σ/ n 10 / 100
p-value = 2P(Z > 0) = 2(.5) = 1.00
248
There is not enough evidence to infer that µ ≠ 100.
11.11 Rejection region: z > z.01= 2.33

x −µ 80 − 70
z= = = 5.00
σ/ n 20 / 100
p-value = p(z > 5.00) = 0
There is enough evidence to infer that µ > 70.
11.12 Rejection region: z < −z.05 = −1.645
x −µ 48 − 50
z= = = −1.33
σ/ n 15 / 100
p-value = P(Z < –1.33) = .0918
There is not enough evidence to infer that µ < 50.
x −µ 52 − 50
11.13a. z = = = 1.20
σ/ n 5/ 9
p-value = P(Z > 1.20) = 1 – .8849 = .1151
249
x −µ 52 − 50
b. z = = = 2.00
σ/ n 5 / 25
p-value = P(Z > 2.00) = .5 – .4772 = .0228.
x −µ 52 − 50
c. z = = = 4.00
σ/ n 5 / 100
p-value = P(Z > 4.00) = 0.
d. The value of the test statistic increases and the p-value decreases.
x −µ 190 − 200
11.14a. z = = = −.60
σ/ n 50 / 9
p-value = P(Z < –.60) = .5 – .2257 = .2743
x −µ 190 − 200
b. z = = = −1.00
σ/ n 30 / 9
p-value = P(Z < –1.00) = .1587
x −µ 190 − 200
c z= = = −3.00
σ/ n 10 / 9
p-value = P(Z < –3.00) = .0013
d. The value of the test statistic decreases and the p-value decreases.
x −µ 21 − 20
11.15 a. z = = = 1.00
σ/ n 5 / 25
p-value = 2P(Z > 1.00) = 2(1 – .8413) = .3174
x −µ 22 − 20
b. z = = = 2.00
σ/ n 5 / 25
p-value = 2P(Z > 2.00) = 2(1 – .9772) = .0456
x −µ 23 − 20
c. z = = = 3.00
σ/ n 5 / 25
p-value = 2P(Z > 3.00) = 2(1 – .9987) = .0026
d. The value of the test statistic increases and the p-value decreases.
x −µ 99 − 100
11.16 a. z = = = −1.25
σ/ n 8 / 100
p-value = 2P(Z < –1.25) = 2(.1056) = .2112
x −µ 99 − 100
b. z = = = −.88
σ/ n 8 / 50
p-value = 2P(Z < –.88) = 2(.1894) = .3788
250
x −µ 99 − 100
c. z = = = −.56
σ/ n 8 / 20
p-value = 2P(Z < –.56) = 2(.2877) = .5754
d. The value of the test statistic increases and the p-value increases.
x −µ 990 − 1000
11.17 a. z = = = −4.00
σ/ n 25 / 100
p-value = P(Z < –4.00) = 0
x −µ 990 − 1000
b. z = = = −2.00
σ/ n 50 / 100
p-value = P(Z < –2.00) = .0228
x −µ 990 − 1000
c. z = = = −1.00
σ/ n 100 / 100
p-value = P(Z < –1.00) = .1587
d. d. The value of the test statistic increases and the p-value increases.
x −µ 72 − 60
11.18 a. z = = = 3.00
σ/ n 20 / 25
p-value = P(Z > 3.00) = 1 – .9987 = .0013
x −µ 68 − 60
b. z = = = 2.00
σ/ n 20 / 25
p-value = P(Z > 2.00) = 1 – .9772 = .0228
x −µ 64 − 60
c. z = = = 1.00
σ/ n 20 / 25
p-value = P(Z > 1.00) = 1 – .8413 = .1587
d. The value of the test statistic decreases and the p-value increases.
x −µ 178 − 170
11.19 a z = = = 1.74
σ/ n 65 / 200
p-valu.e = P(Z > 1.74) = 1 – .9591 = .0409
x −µ 178 − 170
b. z = = = 1.23
σ/ n 65 / 100
p-value = P(Z > 1.23) = 1 – .8907 = .1093
c. The value of the test statistic increases and the p-value decreases.
x −µ 178 − 170
11.20 a z = = = 4.57
σ/ n 35 / 400
251
p-value = P(Z > 4.57) = 0.
x −µ 178 − 170
b z= = = 1.60
σ/ n 100 / 400
p-value = P(Z > 1.60) = 1 – .9452 = .0548
The value of the test statistic decreases and the p-value increases.
11.21 See Table 11.1 in the book.
x −µ 21.63 − 22
11.22 a z = = = −.62
σ/ n 6 / 100
p-value = P(Z < –.62) = .2676
x −µ 21.63 − 22
bz = = = −1.38
σ/ n 6 / 500
p-value = P(Z < –1.38) = .0838
The value of the test statistic decreases and the p-value decreases.
x −µ 21.63 − 22
11.23 a z = = = −1.83
σ/ n 3 / 220
p-value = P(Z < –1.83) = .0336
x −µ 21.63 − 22
bz = = = −.46
σ/ n 12 / 220
p-value = P(Z < –.46) = .3228
The value of the test statistic increases and the p-value increases.
x − 22
11.24 x z= p-value
6 / 220
22.0 0 .5
21.8 –.49 .3121
21.6 –.99 .1611
21.4 –1.48 .0694
21.2 –1.98 .0239
21.0 –2.47 .0068
20.8 –2.97 .0015
20.6 –3.46 0
20.4 –3.96 0
x −µ 17.55 − 17.09
11.25 a z = = = .84
σ/ n 3.87 / 50
p-value = 2P(Z > .84) = 2(1 – .7995) = 2(.2005) = .4010
x −µ 17.55 − 17.09
bz = = = 2.38
σ/ n 3.87 / 400
252
p-value = 2P(Z > 2.38) = 2(1 – .9913) = 2(.0087) = .0174
The value of the test statistic increases and the p-value decreases.
x −µ 17.55 − 17.09
11.26 a z = = = 2.30
σ/ n 2 / 100
p-value = 2P(Z > 2.30) = 2(1 – .9893) = 2(.0107) = .0214
x −µ 17.55 − 17.09
bz= = = .46
σ/ n 10 / 100
p-value = 2P(Z > .46) = 2(1 – .6772) = 2(.3228) = .6456
The value of the test statistic decreases and the p-value increases.
x − 17.09
11.27a x z= p-value
3.87 / 100
15.0 –5.40 0
15.5 –4.11 0
16.0 –2.82 .0048
16.5 –1.52 .1286
17.0 –.23 .8180
17.5 1.06 .2892
18.0 2.35 .0188
18.5 3.64 0
19.0 4.94 0
11.28 H0 : µ = 5
H1 : µ > 5
x −µ 6−5
z= = = 2.11
σ/ n 1.5 / 10
p-value = P(Z > 2.11) = 1 – .9826 = .0174
There is enough evidence to infer that the mean is greater than 5 cases.
11.29 H 0 : µ = 50
H1 : µ > 50
x −µ 59.17 − 50
z= = = 3.89
σ/ n 10 / 18
p-value = P(Z > 3.89) = 0
There is enough evidence to infer that the mean is greater than 50 minutes.
11.30 H 0 : µ = 12
H1 : µ < 12
253
x −µ 11.00 − 12
z= = = −1.29
σ/ n 3 / 15
p-value = P(Z < –1.29) = .0985
There is enough evidence to infer that the average number of golf balls lost is less than 12.
11.31 H 0 : µ = 36
H1 : µ < 36
x −µ 34.25 − 36
z= = = −.76
σ/ n 8 / 12
p-value = P(Z < –.76) = .2236
There is not enough evidence to infer that the average student spent less time than recommended.
11.32 H0 : µ = 6
H1 : µ > 6
x −µ 6.60 − 6
z= = = .95
σ/ n 2 / 10
p-value = P(Z > .95) = 1 – .8289 = .1711
There is not enough evidence to infer that the mean time spent putting on the 18th green is greater
than 6 minutes.
11.33 H 0 : µ = .50
H1 : µ ≠ .50
x −µ .493 − .50
z= = = −.44
σ/ n .05 / 10
p-value = 2P(Z < –.44) = 2(.3300) = .6600
There is not enough evidence to infer that the mean diameter is not .50 inch.
11.34 H 0 : µ = 25
H1 : µ > 25
x −µ 30.22 − 25
z= = = 1.85
σ/ n 12 / 18
p-value = P(Z > 1.85) = 1 – .9678 =.0322
There is not enough evidence to conclude that the manager is correct.
254
11.35 H 0 : µ = 5,000
H1 : µ > 5,000
x −µ 5,065 − 5,000
z= = = 1.62
σ/ n 400 / 100
p-value = P(Z > 1.62) = 1 – .9474 =.0526
There is not enough evidence to conclude that the claim is true.
11.36 H 0 : µ = 30,000
H1 : µ < 30,000
x −µ 29,120 − 30,000
z= = = −2.06
σ/ n 8,000 / 350
p-value = (P(Z < –2.06) = .0197
There is enough evidence to infer that the president is correct
11.37 H 0 : µ = 560
H1 : µ > 560
x −µ 569.0 − 560
z= = = .80
σ/ n 50 / 20
p-value = P(Z > .80) = 1 – .7881 = .2119
There is not enough evidence to conclude that the dean’s claim is true.
11.38a H 0 : µ = 17.85
H1 : µ > 17.85
x −µ 19.13 − 17.85
z= = = 1.65
σ/ n 3.87 / 25
p-value = P(Z > 1.65) = 1 – .9505 = .0495
There is enough evidence to infer that the campaign was successful.
b We must assume that the population standard deviation is unchanged.
11.39 H0 : µ = 0
H1 : µ < 0
x −µ −1.20 − 0
z= = = −1.41
σ/ n 6 / 50
p-value = P(Z < –1.41) = .0793
There is not enough evidence to conclude that the safety equipment is effective.
255
11.40 H 0 : µ = 55
H1 : µ > 55
x −µ 55.80 − 55
z= = = 2.26
σ/ n 5 / 200
p-value = P(Z > 2.26) = 1 – .9881 = .0119
There is not enough evidence to support the officer’s belief.
11.41 H0 : µ = 4
H1 : µ > 4
x −µ 5.04 − 4
z= = = 4.90
σ/ n 1.5 / 50
p-value = P(Z > 4.90) = 0
There is enough evidence to infer that the expert is correct.
11.42 H 0 : µ = 20
H1 : µ < 20
x −µ 19.39 − 20
z= = = −1.22
σ/ n 3 / 36
p-value = P(Z < –1.22) = .1112
There is not enough evidence to infer that the manager is correct.
11.43 H 0 : µ = 100
H1 : µ > 100
x −µ 105.7 − 100
z= = = 2.25
σ/ n 16 / 40
p-value = P(Z > 2.25) = 1 – .9878 = .0122
There is not enough evidence to infer that the site is acceptable.
11.44 H0 : µ = 4
H1 : µ ≠ 4
x −µ 4.84 − 4
z= = = 3.33
σ/ n 2 / 63
p-value = 2P(Z > 3.33) = 0
There is enough evidence to infer that the average Alpine skier does not ski 4 times per year.
256
11.45 H0 : µ = 5
H1 : µ > 5
x −µ 5.64 − 5
z= = = 1.60
σ/ n 2 / 25
p-value = P(Z > 1.60) = 1 – .9452 = .0548
There is enough evidence to infer that the golf professional’s claim is true.
11.46 H 0 : µ = 32
H1 : µ < 32
x −µ 29.92 − 32
z= = = −2.73
σ/ n 8 / 110
p-value = P(Z < –2.73) = 1– .9968 = .0032
There is enough evidence to infer that there has been a decrease in the mean time away from
desks. A type I error occurs when we conclude that the plan decreases the mean time away from
desks when it actually does not. This error is quite expensive. Consequently we demand a low p-
value. The p-value is small enough to infer that there has been a decrease.
11.47 H 0 : µ = 230
H1 : µ > 230
x −µ 231.56 − 230
z= = = 1.56
σ/ n 10 / 100
p-value = P(Z > 1.56) = 1 – .9406 = .0594
There is not enough evidence to infer that Nike is correct.
x −µ x −µ
11.48 Rejection region: > zα / 2 or < −z α / 2
σ/ n σ/ n
x − 200 x − 200
> z .025 = 1.96 or < –1.96
10 / 100 10 / 100
x > 201.96 or x < 198.04
β = P(198.04 < x < 201.96 given µ = 203)

⎛ 198.04 − 203 x − µ 201.96 − 203 ⎞
= P⎜ < < ⎟ = P( –4.96 < z < –1.04) = .1492 – 0 = .1492
⎜ ⎟
⎝ 10 / 100 σ/ n 10 / 100 ⎠
x −µ
11.49 Rejection region: > zα
σ/ n
x − 1000
> z.01 = 2.33
50 / 25
257
x > 1023.3
⎛ x − µ 1023.3 − 1050 ⎞
β = P( x < 1023.3 given µ = 1050) = P⎜⎜ < ⎟⎟ = P(z < –2.67) = .0038
⎝ σ / n 50 / 25 ⎠
x −µ
11.50 Rejection region: < −z α
σ/ n
x − 50
< −z.05 = −1.645
10 / 40
x < 47.40
⎛ x − µ 47.40 − 48 ⎞
β = P( x > 47.40 given µ = 48) = P⎜⎜ > ⎟ = P(z > –.38) = 1 − .3520 = .6480
⎟
⎝ σ / n 10 / 40 ⎠
11.51
Exercise 11.48
Exercise 11.49
258
Exercise 11.50
x −µ
11.52 a. Rejection region: > zα
σ/ n
x − 100
> z .10 = 1.28
20 / 100
x > 102.56
⎛ x − µ 102.56 − 102 ⎞
β = P( x < 102.56 given µ = 102) = P⎜⎜ < ⎟ = P(z < .28) = .6103
⎟
⎝ σ / n 20 / 100 ⎠
x −µ
b. Rejection region: > zα
σ/ n
x − 100
> z .02 = 2.55
20 / 100
x > 104.11
⎛ x − µ 104.11 − 102 ⎞
β = P( x < 104.11 given µ = 102) = P⎜⎜ < ⎟ = P(z < 1.06) = .8554
⎟
⎝ σ / n 20 / 100 ⎠
c. β increases.
x −µ
11.53 a. Rejection region: < −z α
σ/ n
x − 40
< −z .05 = −1.645
5 / 25
x < 38.36
⎛ x − µ 38.36 − 37 ⎞
β = P( x > 38.36 given µ = 37) = P⎜⎜ > ⎟ = P(z > 1.36) = 1 – .9131 = .0869
⎟
⎝ σ / n 5 / 25 ⎠
x −µ
b. Rejection region: < −z α
σ/ n
x − 40
< −z .15 = −1.04
5 / 25
x < 38.96
259
⎛ x − µ 38.96 − 37 ⎞
β = P( x > 38.96 given µ = 37) = P⎜⎜ > ⎟ = P(z > 1.96) = 1 – .9750 = .0250
⎟
⎝ σ / n 5 / 25 ⎠
c. β decreases.
11.54
Exercise 11.52 a
Exercise 11.52 b
260
Exercise 11.53 a
Exercise 11.53 b
x −µ
11.55 a. Rejection region: < −z α
σ/ n
x − 200
< −z .10 = −1.28
30 / 25
x < 192.31
⎛ x − µ 192.31 − 196 ⎞
β = P( x > 192.31 given µ = 196) = P⎜⎜ > ⎟ = P(z > –.62) = 1 − .2676 = .7324
⎟
⎝ σ / n 30 / 25 ⎠
x −µ
b. Rejection region: < −z α
σ/ n
x − 200
< −z .10 = −1.28
30 / 100
x < 196.16
261
⎛ x − µ 196.16 − 196 ⎞
β = P( x > 196.16 given µ = 196) = P⎜⎜ > ⎟ = P(z > .05) = 1 – .5199 = .4801
⎟
⎝ σ / n 30 / 100 ⎠
c. β decreases.
x −µ
11.56 a. Rejection region: > zα
σ/ n
x − 300
> z .05 = 1.645
50 / 81
x > 309.14
⎛ x − µ 309.14 − 310 ⎞
β = P( x < 309.14 given µ = 310) = P⎜⎜ < ⎟ = P(z < –.15) = .4404
⎟
⎝ σ / n 50 / 81 ⎠
x −µ
b. Rejection region: > zα
σ/ n
x − 300
> z .05 = 1.645
50 / 36
x > 313.71
⎛ x − µ 313.71 − 310 ⎞
β = P( x < 313.71 given µ = 310) = P⎜⎜ < ⎟ = P(z < .45) = .6736
⎟
⎝ σ / n 50 / 36 ⎠
c. β increases.
11.57
Exercise 11.55 a
262
Exercise 11.55 b
Exercise 11.56 a
263
Exercise 11.56 b
11.58
264
11.59
11.60 H 0 : µ = 170
H1 : µ < 170
A Type I error occurs when we conclude that the new system is not cost effective when it actually
is. A Type II error occurs when we conclude that the new system is cost effective when it actually
is not.
The test statistic is the same. However, the p-value equals 1 minus the p-value calculated Example
11.1. That is,
p-value = 1 – .0069 = .9931
We conclude that there is no evidence to infer that the mean is less than 170. That is, there is no
evidence to infer that the new system will not be cost effective.
x −µ
11.61 Rejection region: < − zα
σ/ n
x−0
< −z .10 = −1.28
6 / 50
x < –1.09
⎛ x − µ − 1.09 − (−2) ⎞
β = P( x > –1.09 given µ = –2) = P⎜⎜ > ⎟ = P(z > 1.07) = 1 – .8577 = .1423
⎟
⎝ σ / n 6 / 50 ⎠
β can be decreased by increasing α and/or increasing the sample size.
265
x −µ
11.62 Rejection region: < −z α
σ/ n
x − 22
< −z .10 = −1.28
6 / 220
x < 21.48
⎛ x − µ 21.48 − 21 ⎞
β = P( x > 21.48 given µ = 21) = P⎜⎜ > ⎟ = P(z > 1.19) =1 – .8830 = .1170
⎟
⎝ σ / n 6 / 220 ⎠
The company can decide whether the sample size and significance level are appropriate.
x −µ
11.63 Rejection region: > zα
σ/ n
x − 100
> z .01 = 2.33
16 / 40
x > 105.89
⎛ x − µ 105.89 − 104 ⎞
β = P( x < 105.89 given µ = 104) = P⎜⎜ < ⎟ = P(z < .75) = .7734
⎟
⎝ σ / n 16 / 40 ⎠
x −µ
11.64 Rejection region: < − zα
σ/ n
x − 32
< −z .05 = −1.645
8 / 110
x < 30.75
⎛ x − µ 30.75 − 30) ⎞
β = P( x > 30.75 given µ = 30) = P⎜⎜ > ⎟ = P(z > .98) = 1 – .8365 = .1635
⎟
⎝ σ / n 8 / 110 ⎠
β can be decreased by increasing α and/or increasing the sample size.
x −µ
11.65 i Rejection region: < − zα
σ/ n
x − 10
< −z.01 = −2.33
3 / 100
x < 9.30
⎛ x − µ 9.30 − 9 ⎞
β = P( x > 9.30 given µ = 9) = P⎜⎜ > ⎟ = P(z > 1) = 1 – .8413 = .1587
⎟
⎝ σ / n 3 / 100 ⎠
x −µ
ii Rejection region: < − zα
σ/ n
x − 10
< −z.05 = −1.645
3 / 75
266
x < 9.43
⎛ x − µ 9.43 − 9 ⎞
β = P( x > 9.43 given µ = 9) = P⎜⎜ > ⎟ = P(z > 1.24) = 1 – .8925 = .1075
⎟
⎝ σ / n 3 / 75 ⎠
x −µ
iii Rejection region: < − zα
σ/ n
x − 10
< −z.10 = −1.28
3 / 50
x < 9.46
⎛ x − µ 9.46 − 9 ⎞
β = P( x > 9.46 given µ = 9) = P⎜⎜ > ⎟ = P(z > 1.08) = 1 – .8599 = .1401
⎟
⎝ σ / n 3 / 50 ⎠
Plan ii has the lowest probability of a type II error.
11.66 A Type I error occurs when we conclude that the site is feasible when it is not. The
consequence of this decision is to conduct further testing. A Type II error occurs when we do not
conclude that a site is feasible when it actually is. We will do no further testing on this site, and as
a result we will not build on a good site. If there are few other possible sits, this could be an
expensive mistake.
11.67 H 0 : µ = 20
H1 : µ > 25
x −µ
Rejection region: > zα
σ/ n
x − 20
> z .01 = 2.33
8 / 25
x > 23.72
⎛ x − µ 23.72 − 25 ⎞
β = P( x < 23.72 given µ = 25) = P⎜⎜ < ⎟ = P(z < –.80) = .2119
⎟
⎝ σ / n 8 / 25 ⎠
The process can be improved by increasing the sample size.
267
268
Chapter 12
12.3 a x ± t α / 2 s / n = 510 ± 2.064(125/ 25 ) = 510 ± 51.60; LCL = 458.40, UCL = 561.60
b x ± t α / 2 s / n = 510 ± 2.009(125/ 50 ) = 510 ± 35.51; LCL = 474.49, UCL = 545.51
c x ± t α / 2 s / n = 510 ± 1.984(125/ 100 ) = 510 ± 24.80; LCL = 485.20, UCL = 534.80
12.4 a x ± t α / 2 s / n = 1,500 ± 1.984(300/ 100 ) = 1,500 ± 59.52; LCL = 1,440.48,
UCL = 1,559.52
b x ± t α / 2 s / n = 1,500 ± 1.984(200/ 100 ) = 1,500 ± 39.68; LCL = 1,460.32,
UCL = 1,539.68
c x ± t α / 2 s / n = 1,500 ± 1.984(100/ 100 ) = 1,500 ± 19.84; LCL = 1,480.16,
UCL = 1,519.84
12.5 a x ± t α / 2 s / n = 700 ± 1.645(100/ 400 ) = 700 ± 8.23; LCL = 691.77, UCL = 708.23
b x ± t α / 2 s / n = 700 ± 1.96(100/ 400 ) = 700 ± 9.80; LCL = 690.20, UCL = 709.80
a x ± t α / 2 s / n = 700 ± 2.576(100/ 400 ) = 700 ± 12.88; LCL = 687.12, UCL = 712.88
12.6 a x ± t α / 2 s / n = 10 ± 1.984(1/ 100 ) = 10 ± .20; LCL = 9.80, UCL = 10.20
b x ± t α / 2 s / n = 10 ± 1.984(4/ 100 ) = 10 ± .79; LCL = 9.21, UCL = 10.79
c x ± t α / 2 s / n = 10 ± 1.984(10/ 100 ) = 10 ± 1.98; LCL = 8.02, UCL = 11.98
d The interval widens.
12.7 a x ± t α / 2 s / n = 120 ± 2.009(15/ 51 ) = 120 ± 4.22; LCL = 115.78, UCL = 124.22
b x ± t α / 2 s / n = 120 ± 1.676(15/ 51 ) = 120 ± 3.52; LCL = 116.48, UCL = 123.52
c x ± t α / 2 s / n = 120 ± 1.299(15/ 51 ) = 120 ± 2.73; LCL = 117.27, UCL = 122.73
d The interval narrows.
12.8 a x ± t α / 2 s / n = 63 ± 1.990(8/ 81 ) = 63 ± 1.77; LCL = 61.23, UCL = 64.77
269
b x ± t α / 2 s / n = 63 ± 2.000(8/ 64 ) = 63 ± 2.00; LCL = 61.00, UCL = 65.00
c x ± t α / 2 s / n = 63 ± 2.030(8/ 36 ) = 63 ± 2.71; LCL = 60.29, UCL = 65.71
d The interval widens.
12.9 H 0 : µ = 20
H1 : µ > 20
a Rejection region: t > t α, n −1 = t.05,9 = 1.833
x −µ 23 − 20
t= = = 1.05, p-value = .1597. There is not enough evidence to infer that the
s/ n 9 / 10
population mean is greater than 20.
b Rejection region: t > t α, n −1 = t.05, 29 = 1.699
x −µ 23 − 20
t= = = 1.83, p-value = .0391. There is enough evidence to infer that the population
s/ n 9 / 30
mean is greater than 20.
c Rejection region: t > t α,n −1 = t .05, 49 ≈ 1.676
x −µ 23 − 20
t= = = 2.36, p-value = .0112. There is enough evidence to infer that the population
s/ n 9 / 50
mean is greater than 20.
d As the sample size increases the test statistic increases [and the p-value decreases].
12.10 H 0 : µ = 180
H1 : µ ≠ 180
Rejection region: t < − t α / 2,n −1 = − t .025,199 ≈ −1.972 or t > t α / 2,n −1 = t .025,199 = 1.972
x −µ 175 − 180
at = = = −3.21, p-value = .0015. There is enough evidence to infer that the
s/ n 22 / 200
population mean is not equal to 180.
x −µ 175 − 180
b t= = = −1.57, p-value = .1177. There is not enough evidence to infer that the
s/ n 45 / 200
x −µ 175 − 180
c t= = = −1.18, p-value = .2400. There is not enough evidence to infer that the
s/ n 60 / 200
d. As the s increases, the test statistic increases and the p-value increases.
270
12.11 Rejection region: t < − t α,n −1 = − t .05,99 ≈ −1.660
x −µ 145 − 150
at = = = −1.00, p-value = .1599. There is not enough evidence to infer that the
s/ n 50 / 100
population mean is less than 150.
x −µ 140 − 150
bt = = = −2.00, p-value = .0241. There is enough evidence to infer that the
s/ n 50 / 100
x −µ 135 − 150
ct = = = −3.00, p-value = .0017. There is enough evidence to infer that the
s/ n 50 / 100
population mean is less than 150
d The test statistics decreases and the p-value decreases.
12.12 H 0 : µ = 50
H 0 : µ ≠ 50
a Rejection region: t < − t α / 2,n −1 = − t .05,24 = −1.711 or t > t α / 2,n −1 = t .05, 24 = 1.711
x −µ 52 − 50
t= = = .67, p-value = .5113. There is not enough evidence to infer that the
s/ n 15 / 25
b Rejection region: t < − t α / 2,n −1 = − t .05,14 = −1.761or t > t α / 2,n −1 = t .05,14 = 1.761
x −µ 52 − 50
s/ n 15 / 15
c Rejection region: t < − t α / 2,n −1 = − t .05,4 = −2.132 or t > t α / 2,n −1 = t .05, 4 = −2.132
x −µ 52 − 50
s/ n 15 / 5
d The test statistic decreases and the p-value increases.
12.13 Rejection region: t < − t α,n −1 = − t .10, 49 ≈ −1.299
x −µ 585 − 600
at = = = −2.36, p-value = .0112. There is enough evidence to infer that the
s/ n 45 / 50
x −µ 590 − 600
bt = = = −1.57, p-value = .0613. There is enough evidence to infer that the
s/ n 45 / 50
271
x −µ 595 − 600
ct = = = −.79, p-value = .2179. There is not enough evidence to infer that the
s/ n 45 / 50
d The test statistic increases and the p-value increases.
12.14 Rejection region: t > t α,n −1 = t .01,99 ≈ 2.364
x −µ 106 − 100
at = = = 1.71, p-value = .0448. There is not enough evidence to infer that the
s/ n 35 / 100
x −µ 106 − 100
bt= = = 2.40, p-value = .0091. There is enough evidence to infer that the
s/ n 25 / 100
x −µ 106 − 100
ct = = = 4.00, p-value = .0001. There is enough evidence to infer that the
s/ n 15 / 100
population mean is greater than 100
d The test statistic increases and the p-value decreases.
12.15 a x ± t α / 2 s / n = 40 ± 2.365(10/ 8 ) = 40 ± 8.36; LCL = 31.64, UCL = 48.36
b x ± z α / 2 σ / n = 40 ± 1.96(10/ 8 ) = 40 ± 6.93; LCL = 33.07, UCL = 46.93
c The student t distribution is more widely dispersed than the standard normal; thus, z α / 2 is
smaller than t α / 2 .
12.16 a x ± t α / 2 s / n = 175 ± 2.132(30/ 5 ) = 175 ± 28.60; LCL = 146.40, UCL = 203.60
b x ± z α / 2 σ / n = 175 ± 1.645(30/ 5 ) = 175 ± 22.07; LCL = 152.93, UCL = 197.07
c The student t distribution is more widely dispersed than the standard normal; thus, z α / 2 is
smaller than t α / 2 .
12.17 a x ± t α / 2 s / n = 15,500 ± 1.645(9,950/ 1,000 ) = 15,500 ± 517.59; LCL = 14,982.41,
UCL = 16,017.59
b x ± z α / 2 σ / n = 15,500 ± 1.645(9,950/ 1,000 ) = 15,500 ± 517.59; LCL = 14,982.41,
UCL = 16,017.59
c With n = 1,000 the student t distribution with 999 degrees of freedom is almost identical to the
standard normal distribution.
272
12.18 a x ± t α / 2 s / n = 350 ± 2.576(100/ 500 ) = 350 ± 11.52; LCL = 338.48, UCL = 361.52
b x ± z α / 2 σ / n = 350 ± 2.575(100/ 500 ) = 350 ± 11.52; LCL = 338.48, UCL = 361.52
c With n = 500 the student t distribution with 999 degrees of freedom is almost identical to the
12.19 H 0 : µ = 70
H 0 : µ > 70
a Rejection region: t > t α,n −1 = t .05,10 = 1.812
x −µ 74.5 − 70
s/ n 9 / 11
b Rejection region: z > z α = z .05 = 1.645
x −µ 74.5 − 70
z= = = 1.66, p-value = P(Z > 1.66) = 1 – P(Z< 1.66) = 1 – .9515 = .0485. There
σ/ n 9 / 11
is enough evidence to infer that the population mean is greater than 70.
c The Student t distribution is more dispersed than the standard normal.
12.20 H 0 : µ = 110
H 0 : µ < 110
a Rejection region: t < − t α,n −1 = − t .10,9 = –1.383
x −µ 103 − 110
t= = = −1.30, p-value = .1126. There is not enough evidence to infer that the
s/ n 17 / 10
b Rejection region: z < −z α = z .10 = –1.28
x −µ 103 − 110
z= = = −1.30, p-value = P(Z < –1.30) = .0968. There is enough evidence to infer
σ/ n 17 / 10
that the population mean is less than 110.
c The Student t distribution is more dispersed than the standard normal.
12.21 H 0 : µ = 15
H 0 : µ < 15
273
x −µ 14 − 15
t= = = −1.55, p-value = .0608. There is not enough evidence to infer that the
s/ n 25 / 1,500
b Rejection region: z < −z α = −z .05 = –1.645
x −µ 14 − 15
z= = = −1.55, p-value = P(Z < –1.55) = .0606. There is not enough evidence
σ/ n 25 / 1,500
to infer that the population mean is less than 15.
c With n = 1,500 the student t distribution with 1,499 degrees of freedom is almost identical to the
12.22 a Rejection region: t > t α,n −1 = t .05,999 = 1.645
x −µ 405 − 400
s/ n 100 / 1,000
b Rejection region: z > z α = z .05 = 1.645
x −µ 405 − 400
t= = = 1.58, p-value = P( Z > 1.58) = 1 – .9429 = .0571. There is not enough
s/ n 100 / 1,000
evidence to infer that the population mean is less than 15.
c With n = 1,000 the student t distribution with 999 degrees of freedom is almost identical to the
12.23 H0 : µ = 6
H0 : µ < 6
x −µ 5.69 − 6
t= = = −.68, p-value = .2554. There is not enough evidence to support the
s/ n 1.58 / 12
courier’s advertisement.
12.24 x ± t α / 2 s / n = 24,051 ± 2.145(17,386/ 15 ) = 24,051 ± 9,628; LCL = 14,422,
UCL = 33,680
12.25 H 0 : µ = 20
H 0 : µ > 20
Rejection region: t > t α,n −1 = t .05,19 = 1.729
274
x −µ 20.85 − 20
t= = = .56, p-value = .2902. There is not enough evidence to support the
s/ n 6.76 / 20
doctor’s claim.
12.26 H0 : µ = 8
H0 : µ < 8
Rejection region: t < − t α,n −1 = − t.01,17 = −2.567
x −µ 7.91 − 8
t= = = –4.49, p-value = .0002. There is enough evidence to conclude that the
s/ n .085 / 18
average container is mislabeled.
12.27 x ± t α / 2 s / n = 18.13 ± 2.145(9.75/ 15 ) = 18.13 ± 5.40; LCL = 12.73, UCL =23.53
12.28 x ± t α / 2 s / n = 26.67 ± 1.796(16.52/ 12 ) = 26.67 ± 8.56; LCL = 18.11, UCL = 35.23
12.29 x ± t α / 2 s / n = 17.70 ± 2.262(9.08/ 10 ) = 17.70 ± 6.49; LCL = 11.21, UCL =24.19
12.30 H 0 : µ = 10
H 0 : µ < 10
Rejection region: t < − t α,n −1 = − t .10,9 = −1.383
x −µ 7.10 − 10
t= = = −2.45, p-value = .0185. There is enough evidence to infer that the mean
s/ n 3.75 / 10
proportion of returns is less than 10%.
12.31 x ± t α / 2 s / n = 7.15 ± 1.972(1.65/ 200 ) = 7.15 ± .23; LCL = 6.92, UCL = 7.38
12.32 x ± t α / 2 s / n = 4.66 ± 2.576(2.37/ 240 ) = 4.66 ± .39; LCL = 4.27, UCL = 5.05
Total number: LCL = 100 million (4.27) = 427 million, UCL = 100 million (5.05) = 505 million
12.33 x ± t α / 2 s / n =17.00 ± 1.975(4.31/ 162 = 17.00 ± .67; LCL = 16.33, UCL = 17.67
Total number: LCL = 50 million (16.33) = 816.5 million, UCL = 50 million (17.67) = 883.5
million
275
12.34 x ± t α / 2 s / n = 15,137 ± 1.96(5,263/ 306 = 15,137 ± 590; LCL = 14,547, UCL = 15,727
Total credit card debt: LCL = 50 million (14,547) = $727,350 million, UCL = 50 million (15,727)
= $786,350 million
12.35a. x ± t α / 2 s / n = 59.04 ± 1.980(20.62/ 122 ) = 59.04 ± 3.70; LCL = 55.34,
UCL = 62.74
Total spent on other products: LCL = 2800(55.34 = $154,952, UCL = 2800(62.74) = $175,672
12.36 x ± t α / 2 s / n = 2.67 ± 1.973(2.50/ 188 ) = 2.67 ± .36; LCL = 2.31, UCL = 3.03
12.37 x ± t α / 2 s / n = 34.49 ± 1.96(7.82/ 900 ) = 34.49 ± .51; LCL = 33.98, UCL = 35.00
12.38 x ± t α / 2 s / n = 422.36 ± 1.973(122.77/ 176 ) = 422.36 ± 18.26; LCL = 404.10,
UCL = 440.62
Total cost of congestion: LCL = 128 million (404.10) = $51,725 million, UCL = 128 million
(440.62) = $56,399 million
12.39 x ± t α / 2 s / n = 13.94 ± 1.96(2.16/ 212 ) = 13.94 ± .29; LCL = 13.65, UCL = 14.23
Package of 10: LCL = 13.65(10) = 136.5 days, UCL = 14.23(10) = 142.3 days.
12.40 H 0 : µ = 15
H 0 : µ > 15
Rejection region: t > t α,n −1 = t.05,115 ≈ 1.658
x −µ 15.27 − 15
s/ n 5.72 / 116
mean number of commercials is greater than 15.
12.41 x ± t α / 2 s / n = 3.79 ± 1.960(4.25/ 564 ) = 3.79 ± .35; LCL = 3.44, UCL = 4.14
Total number of visits: LCL = 299,157,000(3.44) = 1,029,100,080 UCL = 299,157,000(4.14)

= 1,238,509,980
12.42 H 0 : µ = 85
H 0 : µ > 85
276
Rejection region: t > t α,n −1 = t .05,84 ≈ 1.663
x −µ 89.27 − 85
t= = = 2.28, p-value = .0127. There is enough evidence to infer that an e-
s/ n 17.30 / 85
grocery will be successful.
12.43 x ± t α / 2 s / n = 15.02 ± 1.990(8.31/ 83 ) = 15.02 ± 1.82; LCL = 13.20, UCL = 16.84
12.44 x ± t α / 2 s / n = 96,100 ± 1.960(34,468/ 473 ) = 96,100 ± 3106; LCL = 92,994,
UCL = 99,206
Total amount of debt: LCL = 7 million(92,994) = 650,958 million UCL = 7 million(99,206)
= 694,442 million
12.45 x ± t α / 2 s / n = 1.507 ± 1.645(.640/ 473 ) = 1.507 ± .048; LCL = 1.459, UCL = 1.555
12.46 H 0 : µ = 12
H 0 : µ > 12
t = 20.89, p-value = 0; there is enough evidence to conclude that the average American adult
completed more than 12 years of education.
277
12.47
LCL = 1.40, UCL = 1.49
12.48 H 0 : µ = 40
H 0 : µ > 40
t = 4.80, p-value = 0; there is enough evidence to conclude that the mean number of hours worked
is greater than 40.
12.49
LCL = 6.98, UCL = 7.80
278
12.50
LCL = 2.85, UCL = 3.02
12.51
t = 21.03, p-value = 0; there is enough evidence to conclude that the average American adult
completed more than 12 years of education.
12.52
LCL = 4.80, UCL = 5.12
279
12.53
LCL = 45.29, UCL = 60.27
12.56 H 0 : σ 2 = 300
H1 : σ 2 ≠ 300
a Rejection region: χ 2 < χ12−α / 2,n −1 = χ .2975,99 ≈ 74.2 or χ 2 > χ α2 / 2,n −1 = χ .2025,99 ≈ 130
(n − 1)s 2 (100 − 1)(220)

χ2 = 2
= = 72.60, p-value = .0427. There is enough evidence to infer that
σ 300
the population variance differs from 300.
b Rejection region: χ 2 < χ12−α / 2,n −1 = χ .2975,49 ≈ 32.4 or χ 2 > χ α2 / 2,n −1 = χ .2025,49 ≈ 71.4
(n − 1)s 2 (50 − 1)(220)

χ2 = 2
= = 35.93, p-value = .1643. There is not enough evidence to infer
σ 300
that the population variance differs from 300.
c Decreasing the sample size decreases the test statistic and increases the p-value of the test.
12.57 H 0 : σ 2 = 100
H1 : σ 2 < 100
a Rejection region: χ 2 < χ12−α,n −1 = χ .299,49 ≈ 29.7
(n − 1)s 2 (50 − 1)(80)

χ2 = 2
= = 39.20, p-value = .1596. There is not enough evidence to infer that
σ 100
the population variance is less than 100.
b Rejection region: χ 2 < χ12−α,n −1 = χ .299,99 ≈ 70.1
(n − 1)s 2 (100 − 1)(80)

χ2 = 2
σ 100
the population variance is less than 100.
c Increasing the sample size increases the test statistic and decreases the p-value.
280
( n − 1)s 2 ( n − 1)s 2 (15 − 1)(12)
12.58 a LCL = = = = 7.09
χ α2 / 2, n −1 χ .205,14 23.7
(n − 1)s 2 ( n − 1)s 2 (15 − 1)(12)

UCL = = = 25.57
χ 12− α / 2, n −1 χ .295,14 6.57
( n − 1)s 2 ( n − 1)s 2 (30 − 1)(12)

b LCL = = = = 8.17
χ α2 / 2, n −1 χ .205, 29 42.6
(n − 1)s 2 ( n − 1)s 2 (30 − 1)(12)

UCL = = = = 19.66
χ 12− α / 2, n −1 χ .295, 29 17.7
c Increasing the sample size narrows the interval.
( n − 1)s 2 ( n − 1)s 2 (8 − 1)(.00093)

12.59 LCL = = = = .00046,
χ α2 / 2, n −1 χ .205,7 14.1
(n − 1)s 2 ( n − 1)s 2 (8 − 1)(.00093)

UCL = = = = .00300
χ 12− α / 2, n −1 χ .295,7 2.17
12.60 H 0 : σ 2 = 250
H1 : σ 2 < 250
Rejection region: χ 2 < χ12−α,n −1 = χ .290,9 = 4.17
(n − 1)s 2 (10 − 1)(210.22)

χ2 = 2
σ 250
that the population variance has decreased.
12.61 H 0 : σ 2 = 23
H1 : σ 2 ≠ 23
Rejection region: χ 2 < χ12−α / 2,n −1 = χ .295,7 = 2.17 or χ 2 > χ α2 / 2,n −1 = χ .205,7 = 14.1
(n − 1)s 2 (8 − 1)(16.50)
χ2 = 2
σ 23
the population variance has changed.
( n − 1)s 2 ( n − 1)s 2 (10 − 1)(15.43)

12.62 LCL = = = = 7.31
χ α2 / 2, n −1 χ .2025,9 19.0
281
(n − 1)s 2 ( n − 1)s 2 (10 − 1)(15.43)
UCL = = = = 51.43
χ 12− α / 2, n −1 χ .2975,9 2.70
12.63 a H 0 : σ 2 = 250
H1 : σ 2 ≠ 250
Rejection region: χ 2 < χ12−α / 2,n −1 = χ .2975,24 = 12.4 or χ 2 > χ α2 / 2,n −1 = χ .2025,24 = 39.4
(n − 1)s 2 (25 − 1)(270.58)

χ2 = 2
σ 250
that the population variance is not equal to 250.
b Demand is required to be normally distributed.
c The histogram is approximately bell shaped.
12.64 H 0 : σ 2 = 18
H1 : σ 2 > 18
Rejection region: χ 2 > χ α2 ,n −1 = χ .210,244 = 272.704 (from Excel)
(n − 1)s 2 (245 − 1)(22.56)

χ2 = = = 305.81; p-value = .0044. There is enough evidence to infer
σ2 18
that the population variance is greater than 18.
( n − 1)s 2 ( n − 1)s 2 (90 − 1)(4.72)

12.65 LCL = = = = 3.72
χ α2 / 2, n −1 χ .205,89 113
(n − 1)s 2 ( n − 1)s 2 (90 − 1)(4.72)

UCL = = = = 6.08
χ 12− α / 2, n −1 χ .295,89 69.1
12.66 H 0 : σ 2 = 200
H1 : σ 2 < 200
Rejection region: χ 2 < χ12−α,n −1 = χ .295,99 ≈ 77.9
(n − 1)s 2 (100 − 1)(174.47)

χ2 = 2
= = 86.36; p-value = .1863. There is not enough evidence to
σ 200
infer that the population variance is less than 200. Replace the bulbs as they burn out.
282
( n − 1)s 2 ( n − 1)s 2 (25 − 1)(19.68)
12.67 LCL = = = = 11.99
χ α2 / 2, n −1 χ .2025, 24 39.4
(n − 1)s 2 ( n − 1)s 2 (25 − 1)(19.68)

UCL = = = = 38.09
χ 12− α / 2, n −1 χ .2975, 24 12.4
12.70 a p̂ ± z α / 2 p̂(1 − p̂) / n = .48 ± 1.96 .48(1 − .48) / 500 = .48 ± .0438
b p̂ ± z α / 2 p̂(1 − p̂) / n = .48 ± 1.96 .48(1 − .48) / 200 = .48 ± .0692
c p̂ ± z α / 2 p̂(1 − p̂) / n = .48 ± 1.96 .48(1 − .48) / 1000 = .48 ± .0310
12.71 a p̂ ± z α / 2 p̂(1 − p̂) / n = .50 ± 1.96 .50(1 − .50) / 400 = .50 ± .0490
b p̂ ± z α / 2 p̂(1 − p̂) / n = .33 ± 1.96 .33(1 − .33) / 400 = .33 ± .0461
c p̂ ± z α / 2 p̂(1 − p̂) / n = .10 ± 1.96 .10(1 − .10) / 400 = .10 ± .0294
12.72 H 0 : p = .60
H1 : p > .60
p̂ − p .63 − .60
a z= = = .61, p-value = P(Z > .61) = 1 – .7291 =.2709
p(1 − p) / n .60(1 − .60) / 100
p̂ − p .63 − .60
b z= = = .87, p-value = P(Z > .87) = 1 – .8078 = .1922
p(1 − p) / n .60(1 − .60) / 200
p̂ − p .63 − .60
c z= = = 1.22, p-value = P(Z > 1.22) = 1 – .8888 = .1112
p(1 − p) / n .60(1 − .60) / 400
d The p-value decreases.
p̂ − p .73 − .70
12.73 a z = = = .65, p-value = P(Z > .65) = 1 – .7422 =.2578
p(1 − p) / n .70(1 − .70) / 100
p̂ − p .72 − .70
b z= = = .44, p-value = P(Z > .44) = 1 – .6700 =.3300
p(1 − p) / n .70(1 − .70) / 100
p̂ − p .71 − .70
c z= = = .22, p-value = P(Z > .22) = 1 – .5871 =.4129
p(1 − p) / n .70(1 − .70) / 100
283
d. The z statistic decreases and the p-value increases.
2 2
⎛ z p̂(1 − p̂) ⎞⎟ ⎛ 1.645 .5(1 − .5) ⎞
12.74 n = ⎜ α / 2 = ⎜ ⎟ = 752
⎜ B ⎟ ⎜ .03 ⎟
⎝ ⎠ ⎝ ⎠
12.75a .5 ± .03
b Yes, because the sample size was chosen to produce this interval.
12.76 a p̂ ± z α / 2 p̂(1 − p̂) / n = .75 ± 1.645 .75(1 − .75) / 752 = .75 ± .0260
b The interval is narrower.

c Yes, because the interval estimate is better than specified.
2 2
⎛ z p̂(1 − p̂) ⎞⎟ ⎛ 1.645 .75(1 − .75) ⎞
12.77 n = ⎜ α / 2 = ⎜ ⎟ = 564
⎜ B ⎟ ⎜ .03 ⎟
⎝ ⎠ ⎝ ⎠
12.78a .75 ± .03

b Yes, because the sample size was chosen to produce this interval.
12.79 a p̂ ± z α / 2 p̂(1 − p̂) / n = .92 ± 1.645 .92(1 − .92) / 564 = .92 ± .0188
b The interval is narrower.

c Yes, because the interval estimate is better than specified.
12.80 a p̂ ± z α / 2 p̂(1 − p̂) / n = .5 ± 1.645 .5(1 − .5) / 564 = .5 ± .0346
b The interval is wider.

c No because the interval estimate is wider (worse) than specified.
12.81 p̂ = 259/373 = .69
p̂ ± z α / 2 p̂(1 − p̂) / n = .69 ± 1.96 .69(1 − .69) / 373 = .69 ± .0469; LCL = .6431, UCL = .7369
12.82 H 0 : p = .25
H1 : p < .25
p̂ = 41/200 = .205
284
p̂ − p .205 − .25
z= = = −1.47, p-value = P(Z < –1.47) = .0708. There is enough
p(1 − p) / n .25(1 − .25) / 200
evidence to support the officer’s belief.
12.83 p̂ = 204/314 = .65
p̂ ± z α / 2 p̂(1 − p̂) / n = .65 ± 1.645 .65(1 − .65) / 314 = .65 ± .0443; LCL = .6057, UCL = .6943
12.84 H 0 : p = .92
H1 : p > .92
p̂ = 153/165 = .927
p̂ − p .927 − .92
z= = = .33, p-value = P(Z > .33) = 1 – .6293 =.3707. There is
p(1 − p) / n .92(1 − .92) / 165
not enough evidence to conclude that the airline’s on-time performance has improved.
12.85 p̂ = 97/344 = .28
p̂ ± z α / 2 p̂(1 − p̂) / n = .28 ± 1.96 .28(1 − .28) / 344 = .28 ± .0474; LCL = .2326, UCL = .3274
12.86 p̂ = 68/400 = .17
p̂ ± z α / 2 p̂(1 − p̂) / n = .17 ± 1.96 .17(1 − .17) / 400 = .17 ± .0368; LCL = .1332, UCL = .2068
12.87 LCL = .1332(1,000,000)(3.00) = $399,600, UCL = .2068(1,000,000)(3.00) = $620,400
x+2 1+ 2
12.88 ~
p= = = .0147
n + 4 200 + 4
~
p (1 − ~
p) .0147(1 − .0147)
~
p ± zα/ 2 = .0147 ± 1.96 = .0147 ± .0165; LCL = 0 (increased from
n+4 200 + 4
–.0018), UCL = .0312
x+2 3+ 2
12.89 ~
p= = = .0132
n + 4 374 + 4
~
p (1 − ~
p) .0132(1 − .0132)
~
p ± zα/ 2 = .0132 ± 1.645 = .0132 ± .0097; LCL = .0035,
n+4 374 + 4
UCL = .0229
285
x+2 1+ 2
12.90 ~
p= = = .0077
n + 4 385 + 4
~
p (1 − ~
p) .0077(1 − .0077)
~
p ± zα/ 2 = .0077 ± 2.575 = .0077 ± .0114; LCL = 0 (increased from
n+4 385 + 4
–.0037), UCL = .0191
12.91a. p̂ ± z α / 2 p̂(1 − p̂) / n = .1056 ± 1.96 .1056 (1 − .1056 ) / 521 = .1056 ± .0264; LCL =
.0792, UCL = .1320
12.92 LCL = 75,000(.0792) =5,940, UCL = 75,000(.1320) = 9,900
12.93 p̂ ± z α / 2 p̂(1 − p̂) / n = .1202 ± 1.96 .1202 (1 − .1202 ) / 391 = .1202 ± .0322; LCL =
.0880, UCL = .1524
12.94 H 0 : p = .90
H1 : p < .90
p̂ − p .8644 − .90
z= = = –1.58, p-value = P(Z < –1.58) = .0571. There is not
p(1 − p) / n .90(1 − .90) / 177
enough evidence to infer that the satisfaction rate is less than 90%.
12.95 p̂ ± z α / 2 p̂(1 − p̂) / n = .2333 ± 1.96 .2333 (1 − .2333 ) / 120 = .2333 ± .0757; LCL =
.1576, UCL = .3090
12.96 p̂ ± z α / 2 p̂(1 − p̂) / n = .600 ± 1.96 .600(1 − .600) / 1508 = .600 ± .025; LCL = .575,
UCL = .625
Total number of Canadians who prefer artificial Christmas trees: LCL = 6 million(.575) = 3.45
million, UCL = 6 million (.625) = 3.75 million
12.97a. p̂ ± z α / 2 p̂(1 − p̂) / n = .7840 ± 1.96 .7840 (1 − .7840 ) / 426 = .7840 ± .0391; LCL =
.7449, UCL = .8231
12.98 H 0 : p = .50
H1 : p > .50
286
p̂ − p .57 − .50
z= = = 1.40, p-value = P(Z > 1.40) = 1 – .9192 =.0808. There is
p(1 − p) / n .50(1 − .50) / 100
enough evidence to conclude that more than 50% of all business students would rate the book as
excellent.
12.99 Codes 1, 2, and 3 have been recoded to 5.

H 0 : p = .90
H1 : p > .90
p̂ − p .96 − .90
z= = = 2.00, p-value = P(Z > 2.00) = 1 – .9772 =.0228. There is
p(1 − p) / n .90(1 − .90) / 100
enough evidence to conclude that more than 90% of all business students would rate the book as at
least adequate.
12.100 p̂ ± z α / 2 p̂(1 − p̂) / n = .0490 ± 1.96 .0490 (1 − .0490 ) / 5000 = .0490 ± .0060; LCL =
.0430, UCL = .0550

Total number of television households: LCL = 115 million(.0430) = 4.945 million, UCL = 115
million(.0550) = 6.325 million
12.101 H 0 : p = .2155
H1 : p ± .2155
p̂ − p .2442 − .2155
z= = = 2.25, p-value = 2P(Z > 2.25) = 2(1 – .9878)
p(1 − p) / n .2155(1 − .2155) / 1040
=.0244. There is enough evidence to conclude that the proportion of 4-4-3-2 hands is not equal to
21.55%. The reason: hands are not shuffled sufficiently to make the hands random.
12.102a. p̂ ± z α / 2 p̂(1 − p̂) / n = .2031 ± 1.96 .2031(1 − .2031) / 650 = .2031 ± .0309; LCL =
.1722, UCL = .2340

Number: LCL = 5 million (.1722) = .861 million, UCL = 5 million (.2340) = 1.17 million
287
12.103
LCL = .1315, UCL = .1623
12.104a
LCL = .4780, UCL = .5146
LCL = .0284, UCL = .0448
12.105 H 0 : p = .10
H1 : p > .10
288
z = 2.28, p-value = .0115. There is enough evidence to infer that more than 10% of Americans are
self-employed.
12.106
LCL = .1647, UCL = .1935
12.107 H 0 : p = .50
H1 : p > .50
z = 6.17, p-value = 0. There is enough evidence to infer that in 2008 more Americans saw
themselves as Democrats than Republicans.
289
12.108 H 0 : p = .50
H1 : p > .50
12.109 H 0 : p = .50
H1 : p > .50
z = 3.27, p-value = .0005. There is enough evidence to infer that in 2004 more Americans saw
12.110 H 0 : p = .50
H1 : p > .50
290
12.111 H 0 : p = .50
H1 : p > .50
themselves as conservatives than liberals.
12.112 H 0 : p = .50
H1 : p > .50
291
12.113 H 0 : p = .50
H1 : p > .50
themselves as conservatives than liberals
12.114 H 0 : p = .50
H1 : p > .50
292
12.115 From the GSS datasets between 2002 and 2008 there are more Democrats than
Republicans and more conservatives than liberals.
12.116 H 0 : p = .50
H1 : p > .50
12.117 H 0 : p = .50
H1 : p > .50
293
z = -1.30. There is not enough evidence to infer that in 2004 more Americans saw themselves as
Democrats than Republicans. In fact, there is some evidence to the contrary
12.118 H 0 : p = .50
H1 : p > .50
12.119 H 0 : p = .50
H1 : p > .50
294
12.120
LCL = .1526, UCL = .1874

Total number: LCL = 230,151,000(.1526) = 35,121,043, UCL = 230,151,000(.1874)
= 43,130,297
12.121 H 0 : p = .50
H1 : p < .50
295
z = -13.32, p-value = 0. There is enough evidence to infer that less than 50% of Americans always
vote.
12.122 H 0 : p = .51
H1 : p ≠ .51
z =-.539, p-value = .5898. There is not enough evidence to infer that the survey results differ from
the actual results.
12.123 H 0 : p = .53
H1 : p ≠ .53
z = 2.86, p-value = .0042. There is enough evidence to infer that the survey results differ from the
actual results.
12.124 Codes 3 and 4 were changed to 5
p̂ ± z α / 2 p̂(1 − p̂) / n = .7305 ± 1.96 .7305 (1 − .7305 ) / 475 = .7305 ± .0399; LCL = .6906,
UCL = .7704; Market segment size: LCL = 19,108,000 (.6906) = 13,195,985,

UCL = 19,108,000 (.7704) = 14,720,803
296
12.125 Code 2 was changed to 3.
p̂ ± z α / 2 p̂(1 − p̂) / n = .5313 ± 1.96 .5313(1 − .5313) / 320 = .5313 ± .0547; LCL = .4766,
UCL = .5860;
Market segment size: LCL = 15,517,000 (.4766) = 7,395,402 , UCL = 15,517,000 (.5860) =
9,092,962
12.126a. p̂ ± z α / 2 p̂(1 − p̂) / n = .2919 ± 1.96 .2919 (1 − .2919 ) / 1836 = .2919 ± .0208; LCL =
.2711, UCL = .3127

b LCL = 107,194,000 (.2711) = 29,060,293, UCL = 107,194,000 (.3127) = 33,519,564
12.127 p̂ ± z α / 2 p̂(1 − p̂) / n = .1077 ± 1.96 .1077 (1 − .1077 ) / 325 = .1077 ± .0337; LCL =
.0740, UCL = .1414; Market segment size: LCL = 35.6 million(.0740) = 2.634 million,
UCL = 35.6 million(.1414) = 5.034 million
12.128 p̂ ± z α / 2 p̂(1 − p̂) / n = .1748 ± 1.645 .1748 (1 − .1748 ) / 412 = .1748 ± .0308; LCL =
.1440, UCL = .2056; Number: LCL = 187 million(.1440) = 26.928 million, UCL = 187
12.129 p̂ ± z α / 2 p̂(1 − p̂) / n = .1500 ± 1.96 .1500 (1 − .1500 ) / 340 = .1500 ± .0380; LCL =
.1120, UCL = .1880; Number: LCL = 187 million(.1120) = 20.944 million, UCL = 187
12.130 a H 0 : µ = 30
H1 : µ > 30
A B C D
1 t-Test: Mean
2
3 Costs
4 Mean 31.95
5 Standard Deviation 7.19
6 Hypothesized Mean 30
7 df 124
8 t Stat 3.04
9 P(T<=t) one-tail 0.0015
10 t Critical one-tail 1.6572
11 P(T<=t) two-tail 0.0030
12 t Critical two-tail 1.9793
t = 3.04, p-value = .0015; there is enough evidence to infer that the candidate is correct.
297
A B C D
1 t-Estimate: Mean
2
3 Costs
4 Mean 31.95
6 LCL 30.68
7 UCL 33.23
b LCL = 30.68, UCL = 33.23
c The costs are required to be normally distributed.
12.131 H 0 : µ = 60
H1 : µ < 60
A B C D
1 t-Test: Mean
2
3 Times
4 Mean 57.79
7 df 23
8 t Stat -1.64
11 P(T<=t) two-tail 0.1138
t = –1.64, p-value = .0569. There is not enough evidence to conclude that the supplier’s assertion
is correct.
12.132 H 0 : σ 2 = 17
H1 : σ 2 > 17
A B C D
1 Chi Squared Test: Variance
2
3 Times
4 Sample Variance 27.47
5 Hypothesized Variance 17
6 df 19
7 chi-squared Stat 30.71
8 P (CHI<=chi) one-tail 0.0435
9 chi-squared Critical one tail Left-tail 10.1170
10 Right-tail 30.1435
11 P (CHI<=chi) two-tail 0.0869
12 chi-squared Critical two tail Left-tail 8.9065
298
χ 2 = 30.71, p-value = .0435. There is enough evidence to infer that problems are likely.
12.133
A B
1 z-Estimate: Proportion
2 Resolution
3 Sample Proportion 0.358
4 Observations 215
5 LCL 0.304
6 UCL 0.412
LCL = .304, UCL = .412
12.134 a
A B C D
1 t-Estimate: Mean
2
3 Marks
4 Mean 71.88
6 LCL 69.03
7 UCL 74.73
LCL = 69.03, UCL = 74.73
b H 0 : µ = 68
H1 : µ > 68
A B C D
1 t-Test: Mean
2
3 Marks
4 Mean 71.88
7 df 49
8 t Stat 2.74
11 P(T<=t) two-tail 0.0086
t = 2.74, p-value = .0043; there is enough evidence to infer that students with a calculus
background would perform better in statistics than students with no calculus.
299
12.135
A B C D
1 t-Estimate: Mean
2
3 Points
4 Mean 117.54
6 LCL 108.19
7 UCL 126.89
LCL = 108.19, UCL = 126.89
12.136
A B
2 Insurance
4 Observations 250
5 LCL 0.582
6 UCL 0.682
LCL = .582, UCL = .682
12.137 a
A B C D
1 t-Estimate: Mean
2
3 Times
4 Mean 6.91
6 LCL 6.84
7 UCL 6.98
LCL = 6.84, UCL = 6.98
b The histogram is bell shaped.
c H0 : µ = 7
H1 : µ < 7
300
A B C D
1 t-Test: Mean
2
3 Times
4 Mean 6.91
7 df 74
8 t Stat -3.48
11 P(T<=t) two-tail 0.0008
t = –3.48, p-value = .0004; there is enough evidence to infer that postal workers are spending less
than seven hours doing their jobs.
12.138
A B C D
1 t-Estimate: Mean
2
3 Time
4 Mean 6.35
6 LCL 6.05
7 UCL 6.65
LCL = 6.05, UCL = 6.65
12.139
A B C D
1 t-Estimate: Mean
2
3 Times
4 Mean 5.79
6 LCL 5.11
7 UCL 6.47
LCL = 5.11, UCL = 6.47
12.140
A B
2 Tourist
4 Observations 72
5 LCL 0.558
6 UCL 0.776
301
LCL = .558, UCL = .776
12.141 H0 : σ2 = 4
H1 : σ 2 > 4
A B C D
2
3 Lengths
5 Hypothesized Variance 4
6 df 99
χ 2 = 161.25, p-value = .0001; there is enough evidence to conclude that the number of springs
requiring reworking is unacceptably large.
12.142 H 0 : p = .90
H1 : p < .90
A B C D
1 z-Test: Proportion
2
3 Springs
5 Observations 100
6 Hypothesized Proportion 0.9
7 z Stat -1.33
8 P(Z<=z) one-tail 0.0912
9 z Critical one-tail 1.2816
10 P(Z<=z) two-tail 0.1824
11 z Critical two-tail 1.6449
z = –1.33, p-value = .0912; there is enough evidence to infer that less than 90% of the springs are
the correct length.
302
12.143
A B C D
1 t-Estimate: Mean
2
3 Service
4 Mean 1.10
6 LCL 0.94
7 UCL 1.26
LCL = .94, UCL = 1.26
12.144 a H 0 : µ = 9.8
H1 : µ < 9.8
A B C D
1 t-Test: Mean
2
3 Time
4 Mean 9.16
6 Hypothesized Mean 9.8
7 df 149
8 t Stat -2.97
11 P(T<=t) two-tail 0.0036
t = –2.97, p-value = .0018; there is enough evidence to infer that enclosure of preaddressed
envelopes improves the average speed of payments.
H 0 : σ 2 = (3.2 ) =10.24
2
b
H1 : σ 2 < 10.24
A B C D
2
3 Time
5 Hypothesized Variance 10.24
6 df 149
303
χ 2 = 101.58, p-value = .0011; there is enough evidence to infer that the variability in payment
speeds decreases when a preaddressed envelope is sent.
2 2
⎛ z p̂(1 − p̂) ⎞⎟ ⎛ 2.575 .5(1 − .5) ⎞
12.145 n = ⎜ α / 2 = ⎜ ⎟ = 4144
⎜ B ⎟ ⎜
⎝ .02 ⎟
⎠
⎝ ⎠
12.146
A B
2 Concert
4 Observations 600
5 LCL 0.1245
6 UCL 0.1822
Proportion: LCL = .1245, UCL = .1822
Total: LCL = 400,000(.1245) = 49,800 UCL = 400,000(.1822) = 72,880
12.147 Number of cars:

H 0 : µ = 125
H1 : µ > 125
A B C D
1 t-Test: Mean
2
3 Cars
4 Mean 125.80
7 df 4
8 t Stat 0.46
11 P(T<=t) two-tail 0.6702
t = .46, p-value = .3351; there is not enough evidence to infer that the employee is stealing by
lying about the number of cars.
Amount of time
H 0 : µ = 3.5
H1 : µ > 3.5
304
A B C D
1 t-Test: Mean
2
3 Time
4 Mean 3.61
6 Hypothesized Mean 3.5
7 df 628
8 t Stat 7.00
9 P(T<=t) one-tail 0
11 P(T<=t) two-tail 0
t = 7.00, p-value = 0; there is enough evidence to infer that the employee is stealing by lying about
the amount of time.
12.148 a
A B C D
1 t-Estimate: Mean
2
3 Percent Gain/Loss
4 Mean 12.03
6 LCL -5.54
7 UCL 29.61
LCL = -5.54%, UCL = 29.61%
b. H 0 : µ = 16
H1 : µ < 16
A B C D
1 t-Test: Mean
2
3 Percent Gain/Loss
4 Mean 12.03
7 df 19
8 t Stat -0.47
11 P(T<=t) two-tail 0.6420
t = -.47, p-value = .3210; there is not enough evidence to infer that Mr. Cramer does less well than
the S&P 500 index.
305
12.149
LCL = .0508, UCL = .0740
12.150 H 0 : µ = 14.35
H1 : µ > 14.35
t = .908, p-value = .1823. There is not enough evidence to conclude that the debt service ratio has
increased since 2006.
12.151 H 0 : µ = 17.62
H1 : µ > 17.62
306
t = 4.23, p-value = 0. There is enough evidence to conclude that the financial obligations ratio has
increased since 2005.
12.152 H 0 : µ = 25.97
H1 : µ > 25.97
t = .959, p-value = .1693. There is not enough evidence to conclude that the financial obligations
ratio for renters has increased since 2005.
12.153
Proportion: LCL = .7267, UCL = .7768

Total: LCL = 56,991,448, UCL = 60,920,540
12.154 H 0 : µ = 65.8
H1 : µ > 65.8
307
t = 2.44, p-value = .0083. There is enough evidence to conclude that the percentage of total
compensation for wages and salaries increased between 2007 and 2008.
Case 12.1
95% confidence interval estimate of mean weekly consumption per student:
A B C D
1 t-Estimate: Mean
2
3 Cans
4 Mean 1.316
6 LCL 1.218
7 UCL 1.414
308
Estimated Mean Number of Cans
per Student Revenue Potential Profit Current Profit Net
LCL = 1.218 $2,436,000 $652,600 $616,000 $36,600
UCL = 1.414 $ 2,828,000 $789,800 $616,000 173,800
Pepsi should sign the exclusivity agreement.
Case 12.2
Estimated Mean Number of Cans
per Student Revenue Potential Profit Current Profit Net
LCL = 1.218 $2,436,000 $652,600 $1,089,200 $-436,600
UCL = 1.414 $ 2,828,000 $789,800 $1,363,600 $-573,800
Coke would not sign the exclusivity agreement. Coke is expected to lose from the exclusivity
agreement because they currently have a much larger share of the market and would not gain by
paying for exclusivity.
Case 12.3
a 95% confidence interval estimate of the mean medical costs for each of the four age categories:
A B C D
1 t-Estimate: Mean
2
3 Age:45-64
4 Mean 1830
5 Standard Deviation 749
6 LCL 1784
7 UCL 1877
309
A B C D
1 t-Estimate: Mean
2
3 Age:65-74
4 Mean 4494
6 LCL 4381
7 UCL 4607
A B C D
1 t-Estimate: Mean
2
3 Age:75-84
4 Mean 8074
6 LCL 7876
7 UCL 8272
A B C D
1 t-Estimate: Mean
2
3 Age:85+
4 Mean 15957
6 LCL 15572
7 UCL 16342
b Estimated annual costs for 20011

Estimate of mean Estimate of total (1,000s)
Age category Number (1,000s) LCL UCL LCL UCL
45 to 64 9,718 1,784 1,877 17,336,912 18,240,686
65 to 74 2,644 4,381 4,607 11,583,364 12,180,908
75 to 84 1,600 7,876 8,272 12,601,600 13,235,200
85 and over 639 15,572 16,342 9,950,508 10,442,538
Total 14,601 51,472,384 54,099,332
310
Estimated annual costs for 2016
45 to 64 10,013 1,784 1,877 17,863,192 18,794,401
65 to 74 3,344 4,381 4,607 14,650,064 15,405,808
75 to 84 1,718 7,876 8,272 13,530,968 14,211,296
85 and over 738 15,572 16,342 11,492,136 12,060,396
Total 15,813 57,536,360 60,471,901

45 to 64 10,065 1,784 1,877 17,955,960 18,892,005
65 to 74 3,992 4,381 4,607 17,488,952 18,391,144
75 to 84 2,045 7,876 8,272 16,106,420 16,916,240
85 and over 810 15,572 16,342 12,613,320 13,237,020
Total 16,912 64,164,652 67,436,409

45 to 64 9,996 1,784 1,877 17,832,864 18,762,492
65 to 74 4,511 4,381 4,607 19,762,691 20,782,177
75 to 84 2,627 7,876 8,272 20,690,252 21,730,544
85 and over 909 15,572 16,342 14,154,948 14,854,878
Total 18,043 72,440,755 76,130,091

45 to 64 10,016 1,784 1,877 17,868,544 18,800,032
65 to 74 4,846 4,381 4,607 21,230,326 22,325,522

75 to 84 3,169 7,876 8,272 24,959,044 26,213,968
85 and over 1,121 15,572 16,342 17,456,212 18,319,382
Total 18,043 81,514,126 85,658,904
311
312
Chapter 13
13.5a Equal-variances estimator
⎛ 1 1 ⎞ ⎛ (25 − 1)129 2 + (25 − 1)1312 ⎞⎛ 1

( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (524 – 469) ± 2.009 ⎜ ⎟⎜ + 1 ⎞⎟
⎜ ⎟⎝ 25 25 ⎠
⎝ n1 n 2 ⎠ ⎝ 25 + 25 − 2 ⎠
= 55 ± 73.87
b Equal-variances estimator
⎛ 1 1 ⎞ ⎛ (25 − 1)255 2 + (25 − 1)260 2 ⎞⎛ 1

( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (524 – 469) ± 2.009 ⎜ ⎟⎜ + 1 ⎞⎟
⎜ ⎟⎝ 25 25 ⎠
⎝ n1 n 2 ⎠ ⎝ 25 + 25 − 2 ⎠
= 55 ± 146.33
c The interval widens.
d Equal-variances estimator
⎛ 1 1 ⎞
( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (524 – 469)
⎝ n1 n 2 ⎠
⎛ (100 − 1)129 2 + (100 − 1)1312 ⎞⎛ 1 1 ⎞

± 1.972 ⎜ ⎟⎜ + = 55 ± 36.26
⎜ 100 + 100 − 2 ⎟⎝ 100 100 ⎟⎠
⎝ ⎠
e The interval narrows.
13.6 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
a Equal-variances test statistic
Rejection region: t < − t α / 2,ν = − t .025, 22 = –2.074 or t > t α / 2,ν = t .025, 22 = 2.074
( x 1 − x 2 ) − (µ 1 − µ 2 ) (74 − 71) − 0
t= = = .43, p-value = .6703. There
⎛ 1 1 ⎞ ⎛ (12 − 1)18 + (12 − 1)16 2
2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎜
⎝ 12 + 12 − 2 ⎟⎝ 12 12 ⎠
⎠
is not enough evidence to infer that the population means differ.
b Equal-variances test statistic

313
( x 1 − x 2 ) − (µ 1 − µ 2 ) (74 − 71) − 0
t= = = .04, p-value = .9716.
⎛ 1 1 ⎞ ⎛ (12 − 1)210 2 + (12 − 1)198 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
n
⎝ 1 n 2 ⎠
⎜
⎝ 12 + 12 − 2 ⎟⎝ 12 12 ⎠
⎠
There is not enough evidence to infer that the population means differ.
c The value of the test statistic decreases and the p-value increases.
d Equal-variances test statistic
( x 1 − x 2 ) − (µ 1 − µ 2 ) (74 − 71) − 0
t= = = 1.53, p-value = .1282.
⎛ 1 1 ⎞ ⎛ (150 − 1)18 2 + (150 − 1)16 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
⎝ n 1 n 2 ⎠ ⎜ 150 + 150 − 2 ⎟⎝ 150 150 ⎟⎠
⎝ ⎠
There is not enough evidence to infer that the population means differ.
e The value of the test statistic increases and the p-value decreases.
f Rejection region: t < − t α / 2,ν = − t .025, 22 = –2.074 or t > t α / 2,ν = t .025, 22 = 2.074
( x 1 − x 2 ) − (µ 1 − µ 2 ) (76 − 71) − 0
t= = = .72, p-value = .4796. There
⎛ 1 1 ⎞ ⎛ (12 − 1)18 2 + (12 − 1)16 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎜
⎝ 12 + 12 − 2 ⎟⎝ 12 12 ⎠
⎠
is not enough evidence to infer that the population means differ.
g The value of the test statistic increases and the p-value decreases.
13.7 a Unequal-variances estimator
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 64.8 (rounded to 65, approximated by ν = 70 )
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
⎛ s 2 s 2 ⎞ ⎛ 18 2 7 2 ⎞
( x1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (63 – 60) ± 1.667 ⎜ + ⎟ = 3 ± 4.59
⎜ n n ⎟ ⎜ 50 45 ⎟
⎝ 1 2 ⎠ ⎝ ⎠
b Unequal-variances estimator
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 63.1 (rounded to 63, approximated by ν = 60 )
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
⎛ s 2 s 2 ⎞ ⎛ 412 15 2 ⎞
( x1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (63 – 60) ± 1.671 ⎜ + ⎟ = 3 ± 10.38
⎜ n n ⎟ ⎜ 50 45 ⎟
⎝ 1 2 ⎠ ⎝ ⎠
d Unequal-variances estimator
314
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 131 (approximated by ν = 140 )
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
⎛ s 2 s 2 ⎞ ⎛ 18 2 7 2 ⎞
( x1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (63 – 60) ± 1.656 ⎜ + ⎟ = 3 ± 3.22
⎜ n n ⎟ ⎜ 100 90 ⎟
⎝ 1 2 ⎠ ⎝ ⎠
e The interval narrows.
13.8 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
a Unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 200.4 (rounded to 200)
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t > t α,ν = t.05,200 = 1.653
( x 1 − x 2 ) − (µ1 − µ 2 ) (412 − 405) − 0

t= = = .62, p-value = .2689. There is not enough evidence
⎛ s12 s 22
⎞ ⎛ 128 2 54 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 150 150 ⎟
⎝ ⎠ ⎝ ⎠
to infer that µ1 is greater than µ 2 .
b Unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 223.1 (rounded to 223)
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t > t α,ν = t .05, 223 ≈ 1.645
( x 1 − x 2 ) − (µ1 − µ 2 ) (412 − 405) − 0

t= = = 2.46, p-value = .0074. There is enough evidence to
⎛ s12 s 22 ⎞ ⎛ 312 16 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 150 150 ⎟
⎝ ⎠ ⎝ ⎠
infer that µ1 is greater than µ 2 .
c The value of the test statistic increases and the p-value decreases.
d Unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 25.6 (rounded to 26)
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t > t α,ν = t.05, 26 = 1.706
315
( x 1 − x 2 ) − (µ1 − µ 2 ) (412 − 405) − 0
⎛ s12 s 22 ⎞ ⎛ 128 2 54 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 20 20 ⎟
⎝ ⎠ ⎝ ⎠
e The value of the test statistic decreases and the p-value increases.
f Unequal-variances test statistic
( x 1 − x 2 ) − (µ1 − µ 2 ) (409 − 405) − 0

⎛ s12 s 22 ⎞ ⎛ 128 2 54 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 150 150 ⎟
⎝ ⎠ ⎝ ⎠
g The value of the test statistic decreases and the p-value increases.
13.9 a Equal-variances t-test degrees of freedom = 28, Unequal-variances t-test degrees of

freedom =26.4
b Equal-variances t-test degrees of freedom = 24, Unequal-variances t-test degrees of freedom
= 10.7
c Equal-variances t-test degrees of freedom = 98 , Unequal-variances t-test degrees of freedom
= 91.2
d Equal-variances t-test degrees of freedom = 103, Unequal-variances t-test degrees of freedom
= 78.5
13.10 a In all cases the equal-variances t-test degrees of freedom is greater than the unequal-
variances t-test degrees of freedom.
13.11 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = 1.02, p-value = .9856; use equal-variances test statistic
Rejection region: t < − t α,ν = − t .10,10 = −1.372
( x 1 − x 2 ) − (µ 1 − µ 2 ) (361.50 − 381.83) − 0
t= = = −.43, p-value = .3382. The
⎛ 1 1 ⎞ ⎛ (6 − 1)6767.5 + (6 − 1)6653.4 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 6+6−2 ⎠⎝ 6 6 ⎠
manager should choose to use cameras.
316
13.12 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (5.10 − 7.30) − 0
t= = = −2.04, p-value = .0283.
⎛ 1 1 ⎞ ⎛ (10 − 1)5.88 + (10 − 1)5.79 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
n
⎝ 1 n 2 ⎠ ⎝ 10 + 10 − 2 ⎠⎝ 10 10 ⎠
There is enough evidence to infer that there are fewer errors when the yellow ball is used.
13.13 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (12.62 − 14.20) − 0
t= = = −2,17, p-value = .0213. There is enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ 2.769⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 10 11 ⎠
evidence to infer that the users of the device leave higher tip percentages than users who do not
use the device.
13.14 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t < − t α / 2,ν = − t .05,13 = −1.771 or t > t α / 2,ν = t .05,13 = 1.771
( x 1 − x 2 ) − (µ 1 − µ 2 ) (3,372 − 4,093) − 0
t= = = −1.59, p-value = .1368.
⎛ 1 1 ⎞ ⎛ (9 − 1)755,196 + (6 − 1)725,778 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 9+6−2 ⎠⎝ 9 6 ⎠
There is not enough evidence to infer a difference between the two types of vacation expenses.
13.15 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = .690, p-value = .5486; use equal-variances test statistic
317
( x 1 − x 2 ) − (µ 1 − µ 2 ) (172.00 − 176.75) − 0
t= = = . − 839, p-value = .2053. There is not enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ 192.375⎜ + ⎟
n
⎝ 1 n 2 ⎠ ⎝ 12 12 ⎠
evidence to infer that games take longer to complete this year than 5 years ago.
13.16 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (33.33 − 31.50) − 0
t= = = 1.12, p-value = .2761. There is not enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ 16.1667⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 12 12 ⎠
evidence to infer that the speeds differ.
13.17a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Rejection region: t > t α,ν = t.05,38 ≈ 1.684
( x 1 − x 2 ) − (µ 1 − µ 2 ) (36.93 − 31.36) − 0
t= = = 4.61, p-value = 0. There
⎛ 1 1 ⎛ (15 − 1)4.23 2 + (25 − 1)3.35 2 ⎞⎛ 1
⎟⎜ + 1 ⎞⎟
⎞
s 2p ⎜⎜ + ⎟⎟ ⎜
⎝ n 1 n 2 ⎠ ⎜
⎝ 15 + 25 − 2 ⎟⎝ 15 25 ⎠
⎠
is enough evidence to infer that Tastee is superior.
⎛ 1 1 ⎞
b ( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (36.93 – 31.36)
⎝ n1 n 2 ⎠
⎛ (15 − 1)4.23 2 + (25 − 1)3.35 2 ⎞⎛ 1 1 ⎞

± 2.021 ⎜ ⎟⎜ + ⎟
⎜ 15 + 25 − 2 ⎟⎝ 15 25 ⎠
⎝ ⎠
= 5.57 ± 2.43; LCL = 3.14, UCL = 8.00
c The histograms are somewhat bell shaped. The weight gains may be normally distributed.
13.18 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t < − t α / 2,ν = − t .025,238 = −1.960 or t > t α / 2,ν = t .025, 238 = 1.960
318
( x 1 − x 2 ) − (µ 1 − µ 2 ) (10.01 − 9.12) − 0
t= = = 1.55, p-value =
⎛ 1 1 ⎞ ⎛ (120 − 1)4.43 2 + (120 − 1)4.45 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
n
⎝ 1 n 2 ⎠
⎜ 120 + 120 − 2 ⎟⎝ 120 120 ⎟⎠
⎝ ⎠
.1204. There is not enough evidence to infer that oat bran is different from other cereals in terms
of cholesterol reduction?
13.19 a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Two-tail F test: F = .50, p-value = 0; use unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 449
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t < − t α / 2,ν = − t .025, 449 ≈ −1.960 or t > t α / 2,ν = t .025, 449 ≈ 1.960
( x 1 − x 2 ) − (µ1 − µ 2 ) (58.99 − 52.96) − 0

⎛ s12 s 22 ⎞ ⎛ 30.77 2 43.32 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 250 250 ⎟
⎝ ⎠ ⎝ ⎠
evidence to conclude that a difference in mean listening times exist between the two populations.
⎛ s 2 s 2 ⎞ ⎛ 30.77 2 43.32 2 ⎞
b ( x 1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (58.99 –52.96) ± 1.960 ⎜ + ⎟ = 6.03 ± 6.59;
⎜ n 1 n 2 ⎟ ⎜ 250 250 ⎟⎠
⎝ ⎠ ⎝
LCL = –.56, UCL = 12.62
c The histograms are bell shaped.
13.20 a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (59.81 − 57.40) − 0
t= = = 2.88, p-value =
⎛ 1 1 ⎞ ⎛ (125 − 1)7.02 2 + (159 − 1)6.99 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
⎝ n 1 n 2 ⎠ ⎜ 125 + 159 − 2 ⎟⎝ 125 159 ⎟⎠
⎝ ⎠
.0021. There is enough evidence to infer that the cruise ships are attracting younger customers.
319
⎛ 1 1 ⎞
b ( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (59.81 – 57.40)
⎝ n1 n 2 ⎠
⎛ (125 − 1)7.02 2 + (159 − 1)6.99 2 ⎞⎛ 1 1 ⎞

± 2.576 ⎜ ⎟⎜ + = 2.41 ± 2.16;
⎜ 125 + 159 − 2 ⎟⎝ 125 159 ⎟⎠
⎝ ⎠
LCL = .25, UCL = 4.57
13.21a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t < − t α / 2,ν = − t.025,198 ≈ −1.972 or t > t α / 2,ν = t.025,198 ≈ 1.972
( x 1 − x 2 ) − (µ 1 − µ 2 ) (10.23 − 9.66) − 0
t= = = 1.40, p-value =
⎛ 1 1 ⎞ ⎛ (100 − 1)2.87 2 + (100 − 1)2.90 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
n
⎝ 1 n 2 ⎠
⎜ 100 + 100 − 2 ⎟⎝ 100 100 ⎟⎠
⎝ ⎠
.1640. There is not enough evidence to infer that the distance males and females drive differs.
⎛ 1 1 ⎞
b ( x1 − x 2 ) ± t α / 2 s 2p ⎜⎜ + ⎟⎟ = (10.23 – 9.66)
⎝ n1 n 2 ⎠
⎛ (100 − 1)2.87 2 + (100 − 1)2.90 2 ⎞⎛ 1 1 ⎞

± 1.972 ⎜ ⎟⎜ + = .57 ± .80; LCL = –.23, UCL = 1.37
⎜ 100 + 100 − 2 ⎟⎝ 100 100 ⎟⎠
⎝ ⎠
13.22 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Rejection region: t > t α,ν = t .05,58 ≈ 1.671
( x 1 − x 2 ) − (µ 1 − µ 2 ) (115.50 − 110.20) − 0
t= = = .94, p-value = .1753.
1 1 ⎞ ⎛ (30 − 1)21.69 2 + (30 − 1)21.93 2 ⎞⎛ 1
⎟⎜ + 1 ⎞⎟
⎛
s 2p ⎜⎜ + ⎟⎟ ⎜
n
⎝ 1 n 2 ⎠
⎜
⎝ 30 + 30 − 2 ⎟⎝ 30 30 ⎠
⎠
There is not enough evidence to retain supplier A - switch to supplier B.
13.23 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t < − t α / 2,ν = − t .025,594 ≈ −1.960 or t > t α / 2,ν = t .025,594 ≈ 1.960
320
( x 1 − x 2 ) − (µ 1 − µ 2 ) (5.56 − 5.49) − 0
t= = = .16,
⎛ 1 1 ⎞ ⎛ (306 − 1)5.36 2 + (290 − 1)5.58 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
n
⎝ 1 n 2 ⎠
⎜ 306 + 290 − 2 ⎟⎝ 306 290 ⎟⎠
⎝ ⎠
p-value = .8759. There is no evidence of a difference in job tenures between men and women.
13.24a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Two-tail F test: F = 5.18, p-value = .0019; use unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 33.9 (rounded to 34)
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
( x 1 − x 2 ) − (µ1 − µ 2 ) (70.42 − 56.44) − 0

t= = = 2.94, p-value = .0060. There is enough
⎛ s12 s 22 ⎞ ⎛ 20.54 2 9.03 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 24 16 ⎟⎠
⎝ ⎠ ⎝
evidence to conclude that the two packages differ in the amount of time needed to learn how to use
them.
⎛ s 2 s 2 ⎞ ⎛ 20.54 2 9.03 2 ⎞
b ( x1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (70.42 –56.44) ± 2.030 ⎜ + ⎟ = 13.98 ± 9.67;
⎜ n n ⎟ ⎜ 24 16 ⎟⎠
⎝ 1 2 ⎠ ⎝
LCL = 4.31, UCL = 23.65
c The amount of time is required to be normally distributed.
d The histograms are somewhat bell shaped.
13.25a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 276.5 (rounded to 277)
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t < − t α,ν = − t .01, 277 ≈ −2.326
( x 1 − x 2 ) − (µ1 − µ 2 ) (5.02 − 7.80) − 0

t= = = –11.60, p-value = 0. There is enough evidence
⎛ s12 s 22 ⎞ ⎛ 1.39 2 3.09 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 200 200 ⎟⎠
⎝ ⎠ ⎝
to infer that the amount of time wasted in unsuccessful firms exceeds that of successful firms.
321
⎛ s 2 s 2 ⎞ ⎛ 1.39 2 3.09 2 ⎞
b ( x1 − x 2 ) ± t α / 2 ⎜ 1 + 2 ⎟ = (5.02 – 7.80) ± 1.960 ⎜ + ⎟ = –2.78 ± .47;
⎜ n n ⎟ ⎜ 200 200 ⎟⎠
⎝ 1 2 ⎠ ⎝
LCL = –3.25, UCL = –2.31. Workers in unsuccessful companies waste on average between 2.31
and 3.25 hours per week more than workers in successful companies.
13.26 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (.646 − .601) − 0
t= = = 7.54, p-value = 0.
⎛ 1 1 ⎞ ⎛ (125 − 1).045 2 + (145 − 1).053 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
n
⎝ 1 n 2 ⎠
⎜ 125 + 145 − 2 ⎟⎝ 125 145 ⎟⎠
⎝ ⎠
There is enough evidence to conclude that the reaction time of drivers using cell phones is slower
that for non-cell phone users.
13.27 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
( x1 − x 2 ) − (µ 1 − µ 2 ) (.654 − .662) − 0
t= = = –1.17, p-value = .2444.
⎛ 1 1 ⎞ ⎛ (95 − 1).048 2 + (90 − 1).045 2 ⎞⎛ 1
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + 1 ⎞⎟
⎝ n1 n 2 ⎠ ⎜
⎝ 95 + 90 − 2 ⎟⎝ 95 90 ⎠
⎠
There is not enough evidence to infer that the type of discussion affects reaction times.
13.28 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (6.18 − 5.94) − 0
t= = = .90, p-value = .1858.
⎛ 1 1 ⎞ ⎛ (64 − 1)1.59 2 + (81 − 1)1.612 ⎞⎛ 1
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + 1 ⎞⎟
⎝ n 1 n 2 ⎠ ⎜
⎝ 64 + 81 − 2 ⎟⎝ 64 81 ⎠
⎠
There is not enough evidence to infer that people spend more time researching for a financial
planner than they do for a stock broker.
322
13.29 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = .74, p-value = .0446; use unequal-variances test statistic
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 373
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t < − t α,ν = − t.05,373 ≈ −1.645
( x 1 − x 2 ) − (µ1 − µ 2 ) (63.71 − 66.80) − 0

t= = = –4.69, p-value = 0. There is enough evidence to
⎛ s12 s 22
⎞ ⎛ 5.90 2 6.85 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n 1 n 2 ⎟ ⎜ 173 202 ⎟⎠
⎝ ⎠ ⎝
infer that students without textbooks outperform those with textbooks.
13.30 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t < − t α / 2,ν = − t .025,413 ≈ −1.960 or t > t α / 2,ν = t .025, 413 ≈ 1.960
( x 1 − x 2 ) − (µ 1 − µ 2 ) (149.85 − 154.43) − 0
t= = = –2.05,
⎛ 1 1 ⎞ ⎛ (213 − 1)21.82 2 + (202 − 1)23.64 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
n
⎝ 1 n 2 ⎠
⎜ 213 + 202 − 2 ⎟⎝ 213 202 ⎟⎠
⎝ ⎠
p-value = .0412. There is enough evidence to conclude that there are differences in service times
between the two chains.
13.31 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
( x 1 − x 2 ) − (µ 1 − µ 2 ) (488.4 − 498.1) − 0
t= = = -3.98,
⎛ 1 1 ⎞ ⎛ (124 - 1)19.6 2 + (187 − 1)21.9 2 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ +
⎝ n 1 n 2 ⎠ ⎜ 124 + 187 − 2 ⎟⎝ 124 187 ⎟⎠
⎝ ⎠
p-value = 8.63E-05. There is enough evidence to conclude that there are differences in amounts of
sleep between men and women.
13.32 a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
323
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 190
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t < − t α / 2,ν = − t.025,190 ≈ −1.973 or t > t α / 2,ν = t.025,190 ≈ 1.973
( x1 − x 2 ) − (µ1 − µ 2 ) (130.93 − 126.14) − 0

⎛ s12 s 22 ⎞ ⎛ 31.99 2 26.00 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n n ⎟ ⎜ 100 100 ⎟
⎝ 1 2 ⎠ ⎝ ⎠
evidence to infer that differences exist between the two types of customers.
13.33 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
( x1 − x 2 ) − (µ1 − µ 2 ) (73.60 − 69.20) − 0

t= = = .91, p-value = .1849.
⎛ 1 1 ⎞ ⎛ (20 − 1)15.60 2 + (20 − 1)15.06 2 ⎞⎛ 1
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + 1 ⎞⎟
⎝ n1 n 2 ⎠ ⎜
⎝ 20 + 20 − 2 ⎟⎝ 20 20 ⎠
⎠
There is not enough evidence to infer that the new design tire lasts longer than the existing design
tire.
13.34 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
( x1 − x 2 ) − (µ1 − µ 2 ) 60,245 − 63,563) − 0

t= = = −2.09
⎛ 1 1 ⎞ ⎛ (90 − 1)10,506 2 + (90 − 1)10,7552 ⎞⎛ 1 1
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + ⎞⎟
⎝ n1 n 2 ⎠ ⎜ 90 + 90 − 2 ⎟⎝ 90 90 ⎠
⎝ ⎠
t = −2.09 , p-value = .0189. There is enough evidence to conclude that commission salespeople
outperform fixed-salary salespersons
13.35 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Rejection region: t > t α / 2,ν = t.025, 429 ≈ 1.645and t − t α / 2,ν = − t .025, 429 ≈ −1.645
324
( x1 − x 2 ) − (µ1 − µ 2 ) 633.97 − 661.86) − 0
t= = = −4.58
⎛ 1 1 ⎞ ⎛ (93 − 1)49.45 2 + (338 − 1)52.69 2 ⎞⎛ 1
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜ + 1 ⎞⎟
⎝ n1 n 2 ⎠ ⎜
⎝ 93 + 338 − 2 ⎟⎝ 93 338 ⎠
⎠
t = −4.58, p-value = 0. There is enough evidence to conclude there is a difference in scores
between those who have and those who do not have accidents in a three-year period.
13.36 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
(s12 / n 1 + s 22 / n 2 ) 2
ν= = 222
(s12 / n 1 ) 2 (s 22 / n 2 ) 2
+
n1 −1 n 2 −1
Rejection region: t > t α,ν = t .05,222 ≈ 1.645
( x1 − x 2 ) − (µ1 − µ 2 ) (14.20 − 11.27) − 0

t= = = 6.28, p-value = 0. There is enough evidence to
⎛ s12 s 22 ⎞ ⎛ 2.84 2 4.42 2 ⎞
⎜ + ⎟ ⎜ + ⎟
⎜ n n ⎟ ⎜ 130 130 ⎟⎠
⎝ 1 2 ⎠ ⎝
conclude that bottles of wine with metal caps are perceived to be cheaper.
13.37 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Rejection region: t < − t α,ν = − t .05,641 ≈ −1.6472
( x1 − x 2 ) − (µ1 − µ 2 ) ( 496.9 − 511.3) − 0

t= = = −2.54, p-value =
⎛ 1 1 ⎞ ⎛ (355 − 1)73.8 2 + ( 288 − 1)69.12 ⎞⎛ 1 1 ⎞
s 2p ⎜⎜ + ⎟⎟ ⎜ ⎟⎜
⎝ n1 n 2 ⎠ ⎜ 355 + 288 − 2 ⎟⎝ 355 + 288 ⎟⎠
⎝ ⎠
.0057. There is enough evidence to conclude that SAT scores improved after the change in school
start time.
13.38 Two-tail F test: F = 2.13, p-value = 0; use unequal-variances test statistic.
325
LCL = 13,281.97, UCL = 21,823.01.
13.39 Two-tail F test: F = 1.99, p-value = 0; use unequal-variances test statistic.
LCL = 12,036.17, UCL = 16,857.18.
13.40 Using the CPI annual data, 2008 dollars are adjusted to the 2006 base year. The nominal
differential in salaries is summarized below. Without adjusting for inflation, the mean salary
differential between men and women grew by 21.5% in 2008 compared to the mean salary
differentials in 2006. However, after adjusting for inflation, this differential is reduced to 13.8%.
The data show that men earn significantly higher nominal and real incomes than women, and this
differential has grown larger over the years between 2006 and 2008.
326
13.41 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
t = 1.50, p-value = .0677. There is not enough evidence to conclude that government salaries are
higher than that of the private sector.
13.42 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = .765, p-value = .00321; use unequal-variances test statistic
t = -4.65, p-value = 1.88E-06 ≈ 0. There is enough evidence to conclude that Republicans have
more years of education than do Democrats.
13.43 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
327
Two-tail F test: F = 2.175, p-value = 0; use unequal-variances test statistic
t= 9.84, p-value = 2.26E-22 ≈ 0. There is enough evidence to conclude that men draw higher
incomes than women in 2008 (ANES).
13.44 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
t = 9.20, p-value = 1.15E-19 ≈ 0. There is enough evidence to conclude that men draw higher
incomes than women in 2004 (ANES).
13.45 The data are observational. Experimental data could be produced by randomly assigning
babies to either Tastee or the competitor’s product.
13.46 Assuming that the volunteers were randomly assigned to eat either oat bran or another grain
cereal the data are experimental.
328
13.47a. Since 40 students were selected randomly, but were given the choice as to which software
package to use, the data must be observational.
b. Experimental data could have been derived by selecting the 40 students at random and assigning
either of the software packages at random.
c. Students may choose the software package to which they have prior experience and greater
proficiency. The differences in the amount of time needed to learn how to use each software
package may be a function of the popularity of one software package over the other.
13.48a Let students select the section they wish to attend and compare test results.
b Randomly assign students to either section and compare test results.
13.49 Randomly assign patients with the disease to receive either the new drug or a placebo.
13.50a Randomly select finance and marketing MBA graduates and determine their starting
salaries.
b Randomly assign some MBA students to major in finance and others to major in marketing.
Compare starting salaries after they graduate.
c Better students may be attracted to finance and better students draw higher starting salaries.
13.51a The data are observational because to obtain experimental data would entail randomly
assigning some people to smoke and others not to smoke.
b It is possible that some people smoke because of a genetic defect (Genetics have been associated
with alcoholism.), which may also be linked to lung cancer.
c In our society the experiment described in part a is impossible.
13.52 H0 : µD= 0
H1 : µ D < 0
x D − µD − 4.75 − 0
t= = = −3.22, p-value = .0073. There is enough evidence to infer that the
sD / n D 4.17 / 8
Brand A is better than Brand B.
13.53 H0 : µD= 0
H1 : µ D < 0
329
x D − µD − .175 − 0
t= = = −2.20, p-value = .0320. There is enough evidence to infer that ABS is
sD / nD .225 / 8
better.
13.54 H0 : µD= 0
H1 : µ D > 0
Rejection region: t > t α,ν = t .05,6 = 1.943
xD − µD 1.86 − 0
t= = = 1.98, p-value = .0473. There is enough evidence to infer that the
sD / nD 2.48 / 7
camera reduces the number of red-light runners.
13.55a H0 : µD= 0
H1 : µ D < 0
Rejection region: t < −t α,ν = −t.05,11 = −1.796
xD − µD − 1.00 − 0
t= = = –1.15, p-value = .1375. There is not enough evidence to infer that
sD / nD 3.02 / 12
the new fertilizer is better.

sD 3.02
b xD ± tα / 2 = − 1.00 ± 2.201 = −1.00 ± 1.92 ; LCL = –2.92, UCL = .92
nD 12
c The differences are required to be normally distributed
d No, the histogram is bimodal.
e The data are experimental.
f The experimental design should be independent samples.
13.56 a H 0 : µ D = 0
H1 : µ D > 0
xD − µD 3.08 − 0
t= = = 1.82, p-value = .0484. There is enough evidence to infer that
sD / nD 5.88 / 12
companies with exercise programs have lower medical expenses.

sD 5.88
b xD ± tα / 2 = 3.08 ± 2.201 = 3.08 ± 3.74 ; LCL = –.66, UCL = 6.82
nD 12
c Yes because medical expenses will vary by the month of the year.
330
13.57 H0 : µD= 0
H1 : µ D > 0
x D − µD 12.4 − 0
t= = = 1.53, p-value = .0638. There is not enough evidence to infer that
sD / n D 99.1 / 150
mortgage payments have increases in the past 5 years.
13.58 H0 : µD= 0
H1 : µ D ≠ 0
x D − µD − 1.16 − 0
t= = = –3.70, p-value = .0006. There is enough evidence to infer that
sD / n D 2.22 / 50
waiters and waitresses earn different amounts in tips.
sD 30.63
13.59 a x D ± t α / 2 = 19.75 ± 1.684 = 19.75 ± 8.16 ; LCL = 11.59, UCL = 27.91
nD 40
b H0 : µD= 0
H1 : µ D > 0
x D − µD 19.75 − 0
t= = = 4.08, p-value = .0001. There is enough evidence to conclude that
sD / n D 30.63 / 40
companies that advertise in the Yellow Pages have higher sales than companies that do not.
c The histogram of the differences is bell shaped.
d No, because we expect a great deal of variation between stores.
13.60a H0 : µD= 0
H1 : µ D > 0
xD −µD 57.40 − 0
t= = = 16.92, p-value = 0. There is enough evidence to conclude that
sD / nD 13.14 / 15
heating costs for insulated homes is less than that for uninsulated homes.
sD 13.14
b xD ± tα / 2 = 57.40 ± 2.145 = 57.40 ± 7.28 ; LCL = 50.12, UCL = 64.68
nD 15
331
c Differences are required to be normally distributed.
13.61 H0 : µD= 0
H1 : µ D ≠ 0
x D − µD − 42.94 − 0
t= = = –.91, p-value = .3687. There is not enough evidence to infer men
sD / n D 317.16 / 45
and women spend different amounts on health care.
13.62 H0 : µD= 0
H1 : µ D < 0
Rejection region: t < − t α,ν = − t.05,169 ≈ −1.654
x D − µD − 183.35 − 0
t= = = –1.52, p-value = .0647. There is not enough to infer stock
sD / n D 1568.94 / 170
holdings have decreased.
13.63 H0 : µD= 0
H1 : µ D > 0
xD − µD .0422 − 0
t= = = 1.59, p-value = .0599. There is not enough evidence to conclude
sD / nD .1634 / 38
that ratios are higher this year.
13.64 H0 : µD= 0
H1 : µ D > 0
xD − µD 520.85 − 0
t= = = 2.08, p-value = .0210. There is enough evidence to infer that
sD / nD 1854.92 / 55
company 1’s calculated tax payable is higher than company 2’s.
13.65 H0 : µD= 0
H1 : µ D > 0
332
x D − µD 4.55 − 0
t= = = 2.82, p-value = .0055. There is enough evidence to that the new
sD / n D 7.22 / 20
design tire lasts longer than the existing design.
13.66 The matched pairs experiment reduced the variation caused by different drivers.
13.67 H0 : µD= 0
H1 : µ D > 0
Rejection region: t > t α,ν = t.05, 24 = 1.711
x D − µD 4587 − 0
t= = = 1.00, p-value = .1628. There is not enough evidence to infer that
sD / n D 22,851 / 25
finance majors attract higher salary offers than do marketing majors.
13.68 Salary offers and undergraduate GPA are not as strongly linked as are salary offers and
MBA GPA.
13.69 a H 0 : µ D = 0
H1 : µ D < 0
Rejection region: t < − t α,ν = − t.05, 41 ≈ −1.684
xD − µD − .10 − 0
t= = = –.33, p-value = .3704. There is not enough evidence to infer that for
sD / nD 1.95 / 42
companies where an offspring takes the helm there is a decrease in operating income.
b H0 : µD= 0
H1 : µ D > 0
Rejection region: t > t α,ν = t .05,97 = 1.660
xD − µD 1.24 − 0
t= = = 4.34, p-value = 0. There is enough evidence to infer that when an
sD / nD 2.83 / 98
outsider becomes CEO the operating income increases.
13.70 H0 : µD= 0
H1 : µ D > 0
333
t = 23.35, p-value = 2.99E-103 ≈ 0. There is enough evidence to infer that this generation is more
educated than their fathers.
13.71 H0 : µD= 0
H1 : µ D > 0
t = 26.53, p-value =3.12E-131 ≈ 0. There is enough evidence to infer that this generation is more
educated than their mothers.
13.72 H0 : µD= 0
H1 : µ D > 0
334
t = 2.22, p-value = .0132. There is enough evidence to infer that this generation has more
prestigious occupations than their fathers.
13.73
LCL = 28.18, UCL = 43.38
13.76 a H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 ≠ 1
Rejection region: F > Fα / 2,ν ,ν = F.05,29,29 ≈ 1.88 or
1 2
F < F1−α / 2,ν1 ,ν2 = 1 / Fα / 2,ν2 ,ν1 = 1 / F.05,29,29 ≈ 1 / 1.88 = .53
F = s12 / s 22 = 350/700 =.50, p-value = .0669. There is enough evidence to conclude that the
population variances differ.
335
b Rejection region: F > Fα / 2,ν ,ν = F.025,14,14 = 2.98 or
1 2
F < F1−α / 2,ν1 ,ν2 = 1 / Fα / 2,ν2 ,ν1 = 1 / F.025,14,14 = 1 / 2.98 = .34
F = s12 / s 22 = 350/700 =.50, p-value = .2071. There is not enough evidence to conclude that the
population variances differ.

c The value of the test statistic is unchanged and in this exercise the conclusion changed as well..
⎛ s 2 ⎞ ⎛ 28 ⎞ 1 ⎛ s12
⎟Fα / 2,ν ,ν = ⎛⎜ 28 ⎞⎟4.03 = 5.939
1 ⎞
13.77 a LCL = ⎜ 1 ⎟ = ⎜ ⎟ = .366, UCL = ⎜
⎜ s 2 ⎟ Fα / 2,ν ,ν ⎝ 19 ⎠ 4 . 03 ⎜ s 2 ⎟ 2 1
⎝ 19 ⎠
⎝ 2 ⎠ 1 2 ⎝ 2 ⎠
⎛ s 2 ⎞ ⎛ 28 ⎞ 1 ⎛ s12
⎟Fα / 2,ν ,ν = ⎛⎜ 28 ⎞⎟2.27 = 3.345
1 ⎞
b LCL = ⎜ 1 ⎟ = ⎜ ⎟ = .649, UCL = ⎜
⎜ s 2 ⎟ Fα / 2,ν ,ν ⎝ 19 ⎠ 2.27 ⎜ s 2 ⎟ 2 1
⎝ 19 ⎠
⎝ 2 ⎠ 1 2 ⎝ 2 ⎠
c The interval narrows.
13.78 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 ≠ 1
Rejection region: F > Fα / 2,ν ,ν = F.025,9,10 = 3.78 or
1 2
F < F1−α / 2,ν1 ,ν2 = 1 / Fα / 2,ν2 ,ν1 = 1 / F.025,10,9 = 1 / 3.96 = .25
F = s12 / s 22 = .0000057/.0000114 =.50, p-value = .3179. There is not enough evidence to conclude
that the two machines differ in their consistency of fills.
13.79 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 < 1
Rejection region: F < F1−α,ν ,ν = 1 / Fα,ν ,ν = 1 / F.05,9,9 = 1 / 3.18 = .314
1 2 2 1
F = s12 / s 22 = .1854/.1893 =.98, p-value = .4879. There is not enough evidence to infer that the
second method is more consistent than the first method.
13.80 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 ≠ 1
Rejection region: F > Fα / 2,ν ,ν = F.025,10,10 = 3.72 or
1 2
F < F1−α / 2,ν1 ,ν2 = 1 / Fα / 2,ν2 ,ν1 = 1 / F.025,10,10 = 1 / 3.72 = .269
F = s12 / s 22 = 193.67/60.00 = 3.23, p-value = .0784. There is not enough evidence to infer that the
variances of the marks differ between the two sections.
336
13.81 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 > 1
Rejection region: F > Fα,ν ,ν = F.05,99,99 ≈ 1.39
1 2
F = s12 / s 22 = 19.38/12.70 = 1.53, p-value = .0183. There is enough evidence to infer that limiting
the minimum and maximum speeds reduces the variation in speeds.
13.82 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 ≠ 1
1 2
F < F1−α / 2,ν1 ,ν2 = 1 / Fα / 2,ν2 ,ν1 = 1 / F.025,99,99 ≈ 1 / 1.48 = .68
F = s12 / s 22 = 41,309/19,850 = 2.08, p-value = .0003. There is enough evidence to conclude that the
variances differ.
13.83 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 < 1
Rejection region: F < F1−α,ν ,ν = 1 / Fα,ν ,ν = 1 / F.05,51,51 ≈ 1 / 1.60 = .63
1 2 2 1
F = s12 / s 22 = .0261/.0875 = .298, p-value = 0. There is enough evidence to infer that portfolio 2 is
riskier than portfolio 1.
13.84 H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 ≠ 1
1 2
F < F1−α,ν1 ,ν2 = 1 / Fα,ν2 ,ν1 = 1 / F.05,99,99 ≈ 1 / 1.39 = .72
F = s12 / s 22 = 3.35/10.95 = .31, p-value = 0. There is enough evidence to conclude that the variance
in service times differ.
13.87 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
337
(p̂ 1 − p̂ 2 ) (.45 − .40)
a z= = = .72,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .425(1 − .425)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 100 100 ⎠
p-value = 2P(Z > .72) = 2(1 – .7642) = .4716.

(p̂ 1 − p̂ 2 ) (.45 − .40)
b z= = = 1.43,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .425(1 − .425)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 400 400 ⎠
p-value = 2P(Z > 1.43) = 2(1 – .9236) = .1528.

c The p-value decreases.
13.88 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
(p̂ 1 − p̂ 2 ) (.60 − .55)
a z= = = 1.07,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .575(1 − .575)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 225 225 ⎠
p-value = 2P(Z > 1.07) = 2(1 – .8577) = .2846

(p̂ 1 − p̂ 2 ) (.95 − .90)
bz = = = 2.01,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .925(1 − .925)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 225 225 ⎠
p-value = 2P(Z > 2.01) = 2(1 – .9778) = .0444.

c. The p-value decreases.
(p̂ 1 − p̂ 2 ) (.10 − .05)
d z= = = 2.01,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .075(1 − .075)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 225 225 ⎠
p-value = 2P(Z > 2.01) = 2(1 – .94778) = .0444.

e. The p-value decreases.
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .18(1 − .18) .22(1 − .22)

13.89 a (p̂ 1 − p̂ 2 ) ± z α / 2 + = (.18–.22) ± 1.645 +
n1 n2 100 100
= –.040 ± .0929
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .48(1 − .48) .52(1 − .52)

b (p̂1 − p̂ 2 ) ± z α / 2 + = (.48–.52) ± 1.645 +
n1 n2 100 100
= –.040 ± .1162
338
13.90 H 0 : (p1 − p 2 ) = 0
H1 : ( p 1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.205 − .140)
z= = = 1.70,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .177(1 − .177)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 229 178 ⎠
p-value = P(Z > 1.70) = 1 – .9554 = .0446.

There is enough evidence to conclude that those who paid the regular price are more likely to buy
an extended warranty.
13.91 H 0 : ( p1 − p 2 ) = 0
H1 : ( p 1 − p 2 ) ≠ 0
(p̂ 1 − p̂ 2 ) (.145 − .234)
z= = = −1.69,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .209(1 − .209)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 83 209 ⎠
p-value = 2P(Z < −1.69) = 2(.0455) = .0910.

There is not enough evidence to conclude that new and old accounts are different with respect to
overdue accounts.
13.92 H 0 : (p1 − p 2 ) = 0
H1 : ( p 1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.0196 − .0087)
z= = = 1.74,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .0132(1 − .0132)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 562 804 ⎠
p-value = P(Z > 1.74) = 1 – .9591 = .0409.

There is enough evidence to conclude that those who score under 600 are more likely to default
than those who score 60 or more.
13.93a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
Rejection region: z > z α = z .05 = 1.645
(p̂ 1 − p̂ 2 ) (.56 − .46)

z= = = 4.31, p-value = 0. There is enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .518(1 − .518)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 1100 800 ⎠
evidence to infer that the leader’s popularity has decreased.
339
b H 0 : (p1 − p 2 ) = .05
H1 : (p1 − p 2 ) > .05
(p̂1 − p̂ 2 ) − (p1 − p 2 ) (.56 − .46) − .05

z= = = 2.16,
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .56(1 − .56) .46(1 − .46)
+ +
n1 n2 1100 800
p-value = P(Z > 2.16) = 1 – .9846 = .0154.

There is enough evidence to infer that the leader’s popularity has decreased by more than 5%.
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .56(1 − .56) .46(1 − .46)

c (p̂1 − p̂ 2 ) ± z α / 2 + = (.56 − .46) ± 1.96 +
n1 n2 1100 800
= .10 ± .045
13.94 H 0 : (p1 − p 2 ) = –.08

H1 : (p1 − p 2 ) < –.08
Rejection region: z < −zα = −z.01= –2.33
(p̂1 − p̂ 2 ) − (p1 − p 2 ) (.11 − .28) − (−.08)

z= = = –2.85,
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .11(1 − .11) .28(1 − .28)
+ +
n1 n2 300 300
p-value =P(Z < –2.85) = 1 – .9978 = .0022.

There is enough evidence to conclude that management should adopt process 1.
13.95 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.071 − .064)

z= = = .71,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .068(1 − .068)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 1604 1109 ⎠
p-value = P(Z > .71) = 1 – .7611= .2389.

There is not enough evidence to infer that the claim is false.
13.96 a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
Rejection region: z < −z α = −z .05 = –1.645
340
(p̂ 1 − p̂ 2 ) (.093 − .115)
z= = = –4.04, p-value = 0. There is enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .104(1 − .104)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 6281 6281 ⎠
evidence to infer that Plavix is effective.
13.97 a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
104 189 104 + 189

p̂ 1 = = .0095 p̂ 2 = = .0172 p̂ = = .01335
11,000 11,000 22,000
(p̂ 1 − p̂ 2 ) (.0095 − .0172)
z= = = –4.98, p-value = 0. There is
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .01335(1 − .01335)⎜⎜ + ⎟⎟
⎝ n 1 n 2 ⎠ ⎝ 11,000 11,000 ⎠
enough evidence to infer that aspirin is effective in reducing the incidence of heart attacks.
13.98 H 0 : (p1 − p 2 ) = 0
H1 : ( p 1 − p 2 ) > 0
1,084 997 1,084 + 997

p̂ 1 = = .0985 p̂ 2 = = .0906 p̂ = = .0946
11,000 11,000 22,000
(p̂ 1 − p̂ 2 ) (.0985 − .0906)
z= = = 2.00,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .0946(1 − .0946)⎜⎜ + ⎟⎟
⎝ n 1 n 2 ⎠ ⎝ 11,000 11,000 ⎠
p-value = P(Z > 2.00) = 1 – .9772 = .0228.
There is enough evidence to infer that aspirin leads to more cataracts.
13.99 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
75 132 75 + 132
p̂ 1 = = .0289 p̂ 2 = = .0509 p̂ = = .0399
2,594 2,594 2,594 + 2,594
(p̂ 1 − p̂ 2 ) (.0289 − .0509)
z= = = −4.04, p-value = 0.
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .0399(1 − .0399)⎜⎜ + ⎟⎟
⎝ n 1 n 2 ⎠ ⎝ 2,594 2,594 ⎠
There is enough evidence to infer that Letrozole works.
341
13.100 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
Rejection region: z < −z α = −z .05 = −1.645
88 105 88 + 105
p̂1 = = .2228 p̂ 2 = = .2586 p̂ = = .2409
395 406 395 + 406
(p̂ 1 − p̂ 2 ) (.2228 − .2586)
z= = = −1.19, p-value = P(Z < –1.19) =
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .2409(1 − .2409)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 395 406 ⎠
.1170. There is not enough evidence to infer that exercise training reduces mortality.
13.101 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) ≠ 0
(p̂ 1 − p̂ 2 ) (.9057 − .8878
a z= = = .75,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .8975(1 − .8975)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 350 294 ⎠
p-value = 2P(Z > .75) = 2(1 – .7734)= .4532.

There is not enough evidence to infer that the two populations of car owners differ in their
satisfaction levels.
13.102a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.2632 − .0741)

z= = = 3.35, p-value = 0. There is enough
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .11(1 − .11)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 38 162 ⎠
evidence to conclude that smokers have a higher incidence of heart diseases than nonsmokers.
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 )
b (p̂1 − p̂ 2 ) ± z α / 2 +
n1 n2
.2632(1 − .2632) .0741(1 − .0741)

= (.2632–0741) ± 1.645 + =.1891 ± .1223; LCL = .0668,
38 162
UCL = .3114
13.103 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
342
(p̂ 1 − p̂ 2 ) (.3585 − .3420)

z= = = .53,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .3504(1 − .3504)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 477 462 ⎠
p-value = P(Z > .53) = 1 – .7019 = .2981.

There is not enough evidence to infer that the use of illicit drugs in the United States has increased
in the past decade.
13.104 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
Rejection region: z < −z α / 2 , z > z α / 2 , z.025= 1.96
1 = Success
( p̂1 − p̂ 2 ) (.5169 − .375)
z= = = 4.24, p-value = P(Z > 4.24) = 0
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .4463(1 − .4463)⎜ + ⎟
⎝ 1n n 2 ⎠ ⎝ 445 440 ⎠
There is enough evidence to infer that Canadians and Americans differ in their responses to the
survey question.
13.105 H 0 : (p1 − p 2 ) = –.02

H1 : (p1 − p 2 ) < –.02
(p̂1 − p̂ 2 ) − (p1 − p 2 ) (.055 − .11) − (−.02)

z= = = –1.28,
p̂1 (1 − p̂1 ) p̂ 2 (1 − p̂ 2 ) .055(1 − .055) .11(1 − .11)
+ +
n1 n2 200 200
p-value = P(Z < –1.28) = .1003. There is not enough evidence to choose machine A.
13.106 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
2 = Success
(p̂ 1 − p̂ 2 ) (.8608 − .7875)
z= = = 1.50, p-value = 0.0664
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .8394(1 − .8394)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 194 80 ⎠
There is not enough evidence to infer that those with more education are less likely to work 11
hours or more per day.
343
13.107 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
Rejection region: z < −z α = −z .05= -1.645
(p̂ 1 − p̂ 2 ) (.6058 − .6480)

z= = = -0.66, p-value = .2551
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .6288(1 − .6288)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 104 125 ⎠
There is not enough evidence to infer that Americans are more dissatisfied with their jobs in 2011
than they were in 2008.
13.108 1 = Success
Canada H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.6305 − .5243)

z= = = 2.82, p-value = .0024
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .5834(1 − .5834)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 387 309 ⎠
There is enough evidence to infer that the proportion of Canadians who believe that global
warming is a fact, has fallen in the period between November 2009 and December 2009.
USA H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.4947 − .4613)

z= = = 0.981, p-value = .1634
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .4793(1 − .4793)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 469 401 ⎠
There is not enough evidence to infer that the proportion of Americans who believe that global
Britain H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.4692 − .4300)

z= = = 1.00, p-value = .1587
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .4509(1 − .4509)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 341 300 ⎠
344
There is not enough evidence to infer that the proportion of British who believe that global
13.109 1 = Success
Canada H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.3398 − .2601)

z= = = 2.96, p-value = .0015
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .2990(1 − .2990)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 565 592 ⎠
There is enough evidence to infer that the proportion of Canadians who agree that climate change
is the biggest issue to worry about, has fallen in the period between November 2008 and
November 2009.
USA H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.2604 − .1809)

z= = = 3.30, p-value = .0005
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .2212(1 − .2212)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 603 586 ⎠
There is enough evidence to infer that the proportion of Americans who agree that climate change
is the biggest issue to worry about, has fallen in the period between November 2008 and
November 2009.
Britain H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.2606 − .1509)

z= = = 4.14, p-value = 0
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .2041(1 − .2041)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 449 477 ⎠
There is enough evidence to infer that the proportion of British who agree that climate change is
the biggest issue to worry about, has fallen in the period between November 2008 and November
2009.
345
13.110 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
(p̂ 1 − p̂ 2 ) (.1385 − .0905)

z= = z = 2.04,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .1035(1 − .1035)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 231 619 ⎠
p-value = P(Z > 2.04) = 1 – .9793 = .0207. There is enough evidence to conclude that health
conscious adults are more likely to buy Special X.
13.111a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
68 29 68 + 29
p̂1 = = .4172 p̂ 2 = = .2685 p̂ = = .3579
163 108 163 + 108
(p̂ 1 − p̂ 2 ) (.4172 − .2685)
z= = = 2.50,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .3579(1 − .3579)⎜ + ⎟
⎝ n 1 n 2 ⎠ ⎝ 163 108 ⎠
p-value = P(Z > 2.50) = 1 – .9938 = .0062. There is enough evidence to conclude that members of
segment 1 are more likely to use the service than members of segment 4.
b H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) ≠ 0
Rejection region: z < −z α / 2 = −z .05 = –1.96 or z > z α / 2 = z .05 = 1.96
20 10 20 + 10
p̂1 = = .3704 p̂ 2 = = .4348 p̂ = = .3896
54 23 54 + 23
(p̂ 1 − p̂ 2 ) (.3704 − .4348
z= = = −.53,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .3896(1 − .3896)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 54 23 ⎠
p-value = 2P(Z < –.53) = 2(.2981) = .5962. There is not enough evidence to infer that retired
persons and spouses that work in the home differ in their use of services such as Quik Lube.
13.112 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) ≠ 0
Rejection region: z < −z α / 2 = −z .025 = –1.96 or z > z α / 2 = z .025 = 1.96
346
(p̂ 1 − p̂ 2 ) (.0995 − .1297)
z= = = –1.25,
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
p̂(1 − p̂)⎜⎜ + ⎟⎟ .1132(1 − .1132)⎜ + ⎟
⎝ 1 n n 2 ⎠ ⎝ 382 316 ⎠
p-value = 2P(Z < –1.25) = 2(.1056) = .2112.

There is not enough evidence to infer differences between the two sources.
13.113 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = -.172, p-value = .8638. There is not enough evidence to infer a difference in the proportion of
men and women completing a graduate degree.
13.114 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
.
z = 4.61, p-value = 0. There is enough evidence to infer a difference in the proportion of men and
women who decide to work for themselves.
347
13.115 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = 1.26, p-value = .2084. There is not enough evidence to infer a difference in the proportion of
men and women who answer the question correctly.
13.116 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
men and women who answer the question correctly.
13.117 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
348
z = 2.11, p-value = .0348. There is enough evidence to infer a difference in the proportion of men
and women who answer the question correctly.
13.118 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
.
z = 5.13, p-value = 0. There is enough evidence to infer a difference in the proportion of men and
women who answer the question correctly.
13.119 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
349
workers employed by the government between 2006 and 2008.
13.120 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
.
z = .40, p-value = .6894. There is not enough evidence to infer a difference in the proportion of
those favoring capital punishment between 2006 and 2008.
13.121 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
350
those favoring police permits to buy a gun between 2006 and 2008.
13.122
GSS2002 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = 2.40, p-value = .0164. There is enough evidence to infer a difference in the proportion of those
completing a graduate degree between Democrats and Republicans in 2002.
GSS2004 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
351
z = 0.29, p-value = .7716. In 2004, there is not enough evidence to infer a difference in the
proportion of Democrats versus Republicans who have completed a graduate degree.
GSS2006 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = 2.24, p-value = .0250. In 2006, there is enough evidence to infer a difference in the proportion
of Democrats versus Republicans who have completed a graduate degree, with Democrats
corresponding to the larger proportion.
GSS2008 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
352
z = .99, p-value = .3202. In 2008, there is not enough evidence to infer a difference in the
proportion of Democrats versus Republicans who have completed a graduate degree.
13.123 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = 2.23, p-value = .0258. In 2008, there is enough evidence to infer a difference in the proportion
of Democrats versus Republicans who are employed, with Republicans corresponding to the larger
proportion.
13.124 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
353
z = -3.69, p-value = 0. In 2008, there is enough evidence to infer a difference in the proportion of
Democrats versus Republicans who have health insurance.
13.125 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) ≠ 0
z = -0.63, p-value = .5316. In 2008, there is not enough evidence to infer a difference in the
proportion of Democrats versus Republicans who always vote.
13.126a. H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) > 0
354
z = 2.49, p-value = .0065. There is enough evidence to conclude that there has been a decrease in
participation among boys over the past 10 years.
b. H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) > 0
z = .893, p-value = .1859. There is not enough evidence to conclude that there has been a decrease
in participation among girls over the past 10 years.
c. H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p2 ) > 0
355
z = 2.61, p-value = .0045. There is enough evidence to infer girls are less likely to participate in
sports than boys in 2011.
13.127 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Two-tail F test: F = .859, p-value = .7850; use equal-variances t-test
A B C
1 t-Test: Two-Sample Assuming Equal Variances
2
3 During Before
4 Mean 5746.07 5372.13
5 Variance 167289 194772
6 Observations 15 24
7 Pooled Variance 184373
8 Hypothesized Mean Difference 0
9 df 37
10 t Stat 2.65
11 P(T<=t) one-tail 0.0059
13 P(T<=t) two-tail 0.0119
t = 2.65, p-value = .0059. There is enough evidence to conclude that the campaign is successful.
13.128 Gross sales must increase by 50/.20 = $250 to pay for ads.
H 0 : (µ1 − µ 2 ) = 250
H1 : (µ1 − µ 2 ) > 250
356
A B C
2
3 During Before
4 Mean 5746.07 5372.13
5 Variance 167289 194772
9 df 37
10 t Stat 0.88
11 P(T<=t) one-tail 0.1931
13 P(T<=t) two-tail 0.3862
t = .88, p-value = .1931. There is not enough evidence to conclude that the ads are profitable.
13.129 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = .551, p-value = .0054; use unequal-variances t-test
A B C
1 t-Test: Two-Sample Assuming Unequal Variances
2
3 Group 1 Group 2
4 Mean 4.94 9.48
5 Variance 11.19 20.29
8 df 158
9 t Stat -8.73
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = –8.73, p-value = 0. There is enough evidence to conclude that men and women who suffer
heart attacks vacation less than those who do not suffer heart attacks.
13.130 H0 : µD= 0
H1 : µ D < 0
357
A B C
1 t-Test: Paired Two Sample for Means
2
3 Drug Placebo
4 Mean 18.43 22.03
5 Variance 30.39 66.37
7 Pearson Correlation 0.69
9 df 99
10 t Stat -6.09
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = –6.09, p-value = 0. There is enough evidence to infer that the new drug is effective.
13.131 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
A B C
2
3 Vendor Delivered
4 Mean 19.50 20.03
5 Variance 14.35 14.97
7 Pooled Variance 14.62
9 df 358
10 t Stat -1.29
11 P(T<=t) one-tail 0.0996
13 P(T<=t) two-tail 0.1993
t = –1.29, p-value = .1993. There is not enough evidence of a difference in reading time between
the two groups.
13.132 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
358
A B C D
1 z-Test: Two Proportions
2
3 Last Year This Year
4 Sample Proportions 0.6758 0.7539
6 Hypothesized Difference 0
7 z Stat -2.30
8 P(Z<=z) one tail 0.0106
10 P(Z<=z) two-tail 0.0212
z = –2.30, p-value = .0106. There is enough evidence to infer an increase in seatbelt use.
13.133a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 5 Years This Year
4 Mean 32.42 33.72
5 Variance 36.92 45.52
9 df 398
10 t Stat -2.02
11 P(T<=t) one-tail 0.0218
13 P(T<=t) two-tail 0.0436
t = –2.02, p-value = .0218. There is enough evidence to infer that housing cost a percentage of
total income has increased.
b The histograms are be bell shaped.
13.134a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Two-tail F test: F = 1.10, p-value = .8430; use equal-variances t-test
359
A B C
2
3 Male Female
4 Mean 39.75 49.00
5 Variance 803.88 733.16
9 df 38
10 t Stat -1.06
11 P(T<=t) one-tail 0.1490
13 P(T<=t) two-tail 0.2980
t = –1.06, p-value = .2980. There is not enough evidence to conclude that men and women differ
in the amount of time spent reading magazines.
b H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Low High
4 Mean 33.10 56.84
5 Variance 278.69 1047.81
6 Observations 21.00 19
7 Hypothesized Mean Difference 0.00
8 df 26.00
9 t Stat -2.87
10 P(T<=t) one-tail 0.0040
12 P(T<=t) two-tail 0.0080
t = –2.87, p-value = .0040. There is enough evidence to conclude that high-income individuals
devote more time to reading magazines than do low-income individuals.
13.135a H 0 : µ D = 0
H1 : µ D ≠ 0
360
A B C
2
3 Female Male
4 Mean 55.68 56.40
5 Variance 105.64 116.75
9 df 24
10 t Stat -1.13
11 P(T<=t) one-tail 0.1355
13 P(T<=t) two-tail 0.2710
t = –1.13, p-value = .2710. There is no evidence to infer that gender is a factor.
b A large variation within each gender group was expected.
c The histogram of the differences is somewhat bell shaped.
13.136 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 This Year 3 Years Ago
7 z Stat 2.26
10 P(Z<=z) two-tail 0.0238
z = 2.26, p-value = .0119. There is enough evidence to infer that Americans have become more
distrustful of television and newspaper reporting this year than they were three years ago.
13.137 H 0 : (µ1 − µ 2 ) = 25
H1 : (µ1 − µ 2 ) > 25
361
t = 1.75, p-value = .0433. There is enough evidence to conclude that machine A should be
purchased.
13.138 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) ≠ 0
The totals in columns A through D are 5788, 265, 5154, and 332, respectively.
A B C D E
1 z-Test of the Difference Between Two Proportions (Case 1)
2
3 Sample 1 Sample 2 z Stat -4.28
4 Sample proportion 0.045800 0.064400 P(Z<=z) one-tail 0.0000
5 Sample size 5788 5154 z Critical one-tail 1.6449
6 Alpha 0.05 P(Z<=z) two-tail 0.0000
z = –4.28, p-value = 0. There is enough evidence to infer that the defective rate differs between the
two machines.
13.139 H0 : µD= 0
H1 : µ D > 0
362
Dry Cleaner
A B C
2
3 Dry C Before Dry C After
4 Mean 168.00 165.50
5 Variance 351.38 321.96
9 df 13
10 t Stat 0.96
11 P(T<=t) one-tail 0.1780
13 P(T<=t) two-tail 0.3559
t = .96, p-value = .1780. There is not enough evidence to conclude that the dry cleaner sales have
decreased.
Doughnut shop
A B C
2
3 Donut Before Donut After
4 Mean 308.14 295.29
5 Variance 809.67 812.07
9 df 13
10 t Stat 3.24
11 P(T<=t) one-tail 0.0032
13 P(T<=t) two-tail 0.0065
t = 3.24, p-value = .0032. There is enough evidence to conclude that the doughnut shop sales have
decreased.
363
Convenience store
A B C
2
3 Convenience Before Convenience After
4 Mean 374.64 348.14
5 Variance 2270.40 2941.82
9 df 13
10 t Stat 7.34
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = 7.34, p-value = 0. There is enough evidence to conclude that the convenience store sales have
decreased.
13.140 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
z = -4.53, p-value = 5.43E-06 ≈ 0. There is enough evidence to conclude that Facebook users
maintain lower GPA scores.
13.141a H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
364
A B C D
2
3 Depressed Not Depressed
7 z Stat 2.33
10 P(Z<=z) two-tail 0.0200
z = 2.33, p-value = .0100. There is enough evidence to infer that men who are clinically depressed
are more likely to die from heart diseases.
b No, we cannot establish a causal relationship.
13.142 a H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Exercise Drug
4 Mean 13.52 9.92
5 Variance 5.76 13.16
8 df 42
9 t Stat 4.14
10 P(T<=t) one-tail 0.0001
12 P(T<=t) two-tail 0.0002
t = 4.14, p-value = .0001. There is enough evidence that exercise is more effective than medication
in reducing hypertension.
b
A B C D E F
1 t-Estimate of the Difference Between Two Means (Unequal-Variances)
2
3 Sample 1 Sample 2 Confidence Interval Estimate
4 Mean 13.52 9.92 3.60 ± 1.76
5 Variance 5.76 13.16 Lower confidence limit 1.84
6 Sample size 25 25 Upper confidence limit 5.36
7 Degrees of freedom 41.63
8 Confidence level 0.95
LCL = 1.84, UCL = 5.36
365
13.143 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = 1.93, p-value = .0232; use unequal-variances t-test
A B C
2
3 Group 1 Group 2
4 Mean 7.46 8.46
5 Variance 25.06 12.98
8 df 89
9 t Stat -1.14
10 P(T<=t) one-tail 0.1288
12 P(T<=t) two-tail 0.2575
t = –1.14, p-value = .1288. There is not enough evidence to conclude that people who exercise
moderately more frequently lose weight faster
13.144 H0 : µD= 0
H1 : µ D < 0
A B C
2
3 Group 1 Group 2
4 Mean 7.53 8.57
5 Variance 29.77 43.37
9 df 49
10 t Stat -2.40
11 P(T<=t) one-tail 0.0100
13 P(T<=t) two-tail 0.0201
t = –2.40, p-value = .0100. There is enough evidence to conclude that people who exercise
moderately more frequently lose weight faster
366
13.145 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Special K Other
7 z Stat 2.13
10 P(Z<=z) two-tail 0.0332
z = 2.13, p-value = .0166. There is enough evidence to conclude that Special K buyers like the ad
more than non-buyers.
13.146 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Special K Other
7 z Stat 1.20
10 P(Z<=z) two-tail 0.2282
z = 1.20, p-value = .1141. There is not enough evidence to conclude that Special K buyers are
more likely to think the ad is effective.
13.147 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
367
A B C
2
3 Small space Large space
4 Mean 1245.7 1915.8
5 Variance 23812 65566
8 df 39
9 t Stat -11.21
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = –11.21, p-value = 0. There is enough evidence to infer that students write in such a way as to
fill the allotted space.
13.148 H0 : (µ1 − µ2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Computer No Computer
4 Mean 69,933 48,246
5 Variance 63,359,040 101,588,525
8 df 109
9 t Stat 14.07
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = 14.07, p-value = 0. There is enough evidence to conclude that single-person businesses that use
a PC earn more.
13.149 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
368
A B C D
2
3 New Older
7 z Stat 1.26
10 P(Z<=z) two-tail 0.2074
z = 1.26, p-value = .1037. There is not enough evidence to conclude that the new company is
better.
13.150 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Supplement Placebo
4 Mean 19.02 21.85
5 Variance 41.34 25.49
9 df 94
10 t Stat -2.40
11 P(T<=t) one-tail 0.0092
13 P(T<=t) two-tail 0.0183
t = –2.40, p-value = .0092. There is enough evidence to infer that taking vitamin and mineral
supplements daily increases the body's immune system?
13.151 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
369
A B C D
2
3 $100-Limit $3000-Limit
7 z Stat -1.00
10 P(Z<=z) two-tail 0.3196
z = –1.00, p-value = .1598. There is not enough evidence to infer that the dealer at the more
expensive table is cheating.
13.152 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
t = 0.71, p-value = .4763. There is not enough evidence to conclude that math test scores differ
between males and females.
13.153 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
370
A B C
2
3 Female Male
4 Mean 7.27 7.02
5 Variance 2.57 2.85
9 df 198
10 t Stat 1.08
11 P(T<=t) one-tail 0.1410
13 P(T<=t) two-tail 0.2820
t = 1.08, p-value = .2820. There is not enough evidence to conclude that female and male high
school students differ in the amount of time spent at part-time jobs.
13.154 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 City Suburb
4 Mean 2.42 1.97
9 df 146
10 t Stat 2.85
11 P(T<=t) one-tail 0.0025
13 P(T<=t) two-tail 0.0050
t = 2.85, p-value = .0025. There is enough evidence to infer that city households discard more
newspaper than do suburban households.
13.155 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Two-tail F test: F = 2.73, p-value = 0; use unequal-variances t-test
371
A B C
2
3 Teenagers 20-to-30
4 Mean 18.18 14.30
5 Variance 357.32 130.79
8 df 293
9 t Stat 2.28
10 P(T<=t) one-tail 0.0115
12 P(T<=t) two-tail 0.0230
t = 2.28, p-value = .0115.There is enough evidence to infer that teenagers see more movies than do
twenty to thirty year olds.
13.156 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
A B C D
2
3 No HS HS
7 z Stat -3.54
10 P(Z<=z) two-tail 0.0004
z = –3.54, p-value = .0002. There is enough evidence to conclude that Californians who did not
complete high school are less likely to take a course in the university’s evening program.
13.157 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Two-tail F test: F = .433, p-value = 0; use unequal-variances t-test
372
A B C
2
3 Group 1 Groups 2-4
4 Mean 11.58 10.60
5 Variance 9.28 21.41
8 df 644
9 t Stat 4.15
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = 4.15, p-value = 0. There is enough evidence to conclude that on average the market segment
concerned about eating healthy food (group 1) outspends the other market segments.
13.158 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = .501, p-value = 0; use unequal-variances t-test
A B C
2
3 Sale-CDs Sale-fax
4 Mean 59.04 65.57
5 Variance 425.4 849.7
8 df 256
9 t Stat -2.13
10 P(T<=t) one-tail 0.0171
12 P(T<=t) two-tail 0.0341
t = –2.13, p-value = .0171. There is enough evidence to conclude that those who buy the
fax/copier outspend those who buy the package of CD-ROMS.
13.159 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
373
A B C
2
3 No Yes
4 Mean 91,467 97,836
5 Variance 461,917,705 401,930,840
7 Pooled Variance 455,676,297
9 df 519
10 t Stat -2.09
11 P(T<=t) one-tail 0.0184
13 P(T<=t) two-tail 0.0369
t = –2.09, p-value = .0184. There is enough evidence to infer that professors aged 55 to 64 who
plan to retire early have higher salaries than those who don’t plan to retire early.
13.160 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) ≠ 0
The data must first be unstacked. Success = 2
A B C D
2
3 Female Male
7 z Stat -0.45
10 P(Z<=z) two-tail 0.6512
z = −.45, p-value = .6512. There is not enough evidence to conclude that men and women differ in
their choices of Christmas trees.
374
Appendix 13
A13.1 Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 This Year 3 Years Ago
4 Mean 8.29 10.36
9 df 198
10 t Stat -5.09
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = –5.09, p-value = 0. There is overwhelming evidence to conclude that there has been a decrease
over the past three years.
A13.2 a z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
A B C D E
2
3 Sample 1 Sample 2 z Stat 2.83
z = 2.83, p-value = .0024. There is enough evidence to infer that customers who see the ad are
more likely to make a purchase than those who do not see the ad.
b Equal-variances t-test of µ1 − µ 2
375
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Ad No Ad
4 Mean 97.38 92.01
5 Variance 621.97 283.26
9 df 68
10 t Stat 0.90
11 P(T<=t) one-tail 0.1853
13 P(T<=t) two-tail 0.3705
t = .90, p-value = .1853. There is not enough evidence to infer that customers who see the ad and
make a purchase spend more than those who do not see the ad and make a purchase.
c z-estimator of p
A B C D E
1 z-Estimate of a Proportion
2
3 Sample proportion 0.4336 Confidence Interval Estimate
4 Sample size 113 0.4336 ± 0.0914
5 Confidence level 0.95 Lower confidence limit 0.3423
6 Upper confidence limit 0.5250
We estimate that between 34.23% and 52.50% of all customers who see the ad will make a
purchase.
d t-estimator of µ
A B C D
1 t-Estimate: Mean
2
3 Ad
4 Mean 97.38
6 LCL 90.22
7 UCL 104.55
376
We estimate that the mean amount spent by customers who see the ad and make a purchase lies
between $90.22 and $104.55.
A13.3 t-test of µ D
H0 : µD = 0
H1 : µ D > 0
A B C
2
3 Before After
4 Mean 381.00 373.12
5 Variance 39001 40663
9 df 24
10 t Stat 0.70
11 P(T<=t) one-tail 0.2438
13 P(T<=t) two-tail 0.4876
t = .70, p-value = .2438. There is not enough evidence to conclude that the equipment is effective.
A13.4 Frequency of accidents: z -test of p1 − p 2 (case 1)
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D E
2
3 Sample 1 Sample 2 z Stat 0.47
z = .47, p-value = .32053. There is not enough evidence to infer that ABS-equipped cars have
fewer accidents than cars without ABS.
Severity of accidents Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
377
A B C
2
3 No ABS ABS
4 Mean 2075 1714
5 Variance 450,343 390,409
7 Pooled Variance 421,913
9 df 78
10 t Stat 2
11 P(T<=t) one-tail 0.0077
13 P(T<=t) two-tail 0.0153
Estimate of the difference between two means (equal-variances)
A B C D E F
1 t-Estimate of the Difference Between Two Means (Equal-Variances)
2
3 Sample 1 Sample 2 Confidence Interval Estimate
4 Mean 2075 1714 360.48 ± 290
5 Variance 450,343 390,409 Lower confidence limit 71
6 Sample size 42 38 Upper confidence limit 650
7 Pooled Variance 421,913
8 Confidence level 0.95
We estimate that the mean repair cost for non-ABS-equipped cars will be between $71 and $650
more than the mean repair cost for ABS-equipped cars.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
378
A B C
2
3 Discount No Discount
4 Mean 13.06 18.22
5 Variance 30.26 38.13
9 df 98
10 t Stat -4.41
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = –4.41, p-value = 0. There is enough evidence to infer that the discount plan works.
A13.6 Speeds: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Speeds Before Speeds After
4 Mean 31.74 31.42
9 df 198
10 t Stat 1.07
11 P(T<=t) one-tail 0.1424
13 P(T<=t) two-tail 0.2849
t = 1.07, p-value = .1424. There is not enough evidence to infer that speed bumps reduce speeds.
Proper stops: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F =.99, p-value = .9784; use equal-variances t-test
379
A B C
2
3 Stops Before Stops After
4 Mean 7.82 7.98
9 df 198
10 t Stat -0.84
11 P(T<=t) one-tail 0.2021
13 P(T<=t) two-tail 0.4042
t = –.84, p-value = .2021. There is not enough evidence to infer that speed bumps increase the
number of proper stops.
A13.7 t-estimator of µ
A B C D
1 t-Estimate: Mean
2
3 PSI
4 Mean 4.20
6 LCL 3.93
7 UCL 4.46
LCL = 3.93, UCL = 4.46. We estimate that on average tires are between 3.93 and 4.46 pounds per
square inch below the recommended amount.
Tire life: LCL = 100(3.93) = 393, UCL = 100(4.46) = 446. We estimate that the average tire life is
decreased by between 393 and 446 miles.
Gasoline consumption: LCL = .1(3.93) = .393, UCL = .1(4.46) = .446. We estimate that average
gasoline consumption increases by between .393 and .446 gallons per mile.
H0 : µD = 0
H1 : µ D > 0
380
A B C
2
3 Before After
4 Mean 28.94 26.22
5 Variance 61.45 104.30
9 df 49
10 t Stat 3.73
13 P(T<=t) two-tail 0.000
t = 3.73, p-value = .0002. There is enough evidence to infer that the law discourages bicycle use.
A13.9 z -test of p1 − p 2 (case 1)
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Cardizem Placebo
7 z Stat 0.22
10 P(Z<=z) two-tail 0.8252
z = .22, p-value = .4126. There is not enough evidence to indicate that Cardizem users are more
likely to suffer headache and dizziness side effects than non-users.
A13.10 t-test of µ
H 0 : µ = 200
H 1 : µ > 200
381
A B C D
1 t-Test: Mean
2
3 Pedestrians
4 Mean 209.13
7 df 39
8 t Stat 0.96
11 P(T<=t) two-tail 0.3422
t = .96, p-value = .1711. There is not enough evidence to infer that the franchiser should build on
this site.
A13.11 Unequal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Quit Did not quit
4 Mean 2.04 0.72
8 df 316
9 t Stat 14.06
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = 14.06, p-value = 0. There is enough evidence to infer that quitting smoking results in weight
gains.
A13.12 F-test of σ12 / σ 22
H 0 : σ12 / σ 22 = 1
H1 : σ12 / σ 22 > 1
382
A B C
1 F-Test Two-Sample for Variances
2
3 Brand A Brand B
4 Mean 145.95 144.78
5 Variance 16.45 4.25
7 df 99 99
8 F 3.87
9 P(F<=f) one-tail 0.0000
10 F Critical one-tail 1.3941
F = 3.87, p-value = 0. There is overwhelming evidence to infer that Brand B is superior to Brand
A.
A13.13 a z-test of p1 − p 2 (case 1) (Success = 2)
H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
A B C D
2
3 Male most Male less
7 z Stat -2.36
10 P(Z<=z) two-tail 0.0184
z = –2.36, p-value = .0184. There is enough evidence that men’s choices are affected by the
attractiveness of women’s pictures
b z-test of p1 − p 2 (case 1) (Success = 2)
H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
383
A B C D
2
3 Female most Female less
7 z Stat -0.50
10 P(Z<=z) two-tail 0.6188
z = –.50, p-value = .6188. There is not enough evidence to infer that women’s choices are affected
by the attractiveness of men’s pictures.
A13.14 z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
A B C D
2
3 Exercisers Watchers
7 z Stat 1.66
10 P(Z<=z) two-tail 0.0964
z = 1.66, p-value = .0482. There is evidence to infer that exercisers are more likely to remember
the sponsor's brand name than those who only watch.
A13.15 a z-test of p
H 0 : p = 104,320/425,000 = .245
H1 : p > .245
384
A B C D
2
3 Deliver
5 Observations 400
7 z Stat 1.74
10 P(Z<=z) two-tail 0.0812
z = 1.74, p-value = .0406. There is enough evidence to indicate that the campaign will increase
home delivery sales.
b z-test of p
H 0 : p = 110,000/425,000 = .259
H1 : p > .259
A B C D
2
3 Deliver
5 Observations 400
7 z Stat 1.07
10 P(Z<=z) two-tail 0.2834
z = 1.07, p-value = .1417. There is not enough evidence to conclude that the campaign will be
successful.
A13.16 t-tests of µ D
a H 0 : µ D = 40
H1 : µ D > 40
385
A B C
2
3 SAT after SAT before
4 Mean 1235 1162
5 Variance 37970 28844
9 df 39
10 t Stat 2.98
11 P(T<=t) one-tail 0.0024
13 P(T<=t) two-tail 0.0049
t = 2.98, p-value = .0024. There is enough evidence to conclude that the ETS claim is false.
b H 0 : µ D = 110
H1 : µ D < 110
A B C
2
3 SAT after SAT before
4 Mean 1235 1162
5 Variance 37970 28844
9 df 39
10 t Stat -3.39
11 P(T<=t) one-tail 0.0008
13 P(T<=t) two-tail 0.0016
t = –3.39, p-value = .0008. There is enough evidence to conclude that the Kaplan claim is also
false.
A13.17a t-test of µ
H 0 : µ = 1,000
H1 : µ < 1,000
386
A B C D
1 t-Test: Mean
2
3 Potatoes
4 Mean 985.3
7 df 49
8 t Stat -2.11
11 P(T<=t) two-tail 0.0396
T = −2.11; p - value = .0198. There is enough evidence to infer that the supplier is cheating him.
b t-estimator of µ
A B C D
1 t-Estimate: Mean
2
3 Potatoes
4 Mean 985.3
6 LCL 971.3
7 UCL 999.3
Estimate of the total

LCL = 15,000 (971.3) = 14,569,500
UCL = 15,000(999.3) = 14,989,500
A13.18
387
t= -.155, p-value = .4385. There is not enough evidence …
A13.19 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
z = -5.46, p-value = 0. There is enough evidence to conclude that Democrats and Republicans
differ in their correct answer to this question.
A19.20 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
388
z = -1.61, p-value = .0534. There is not enough evidence to conclude that Democrats and
Republicans differ in their correct answer to this question.
A13.21 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
z = -2.01, p-value = .0221. There is enough evidence to conclude that Democrats and Republicans
A13.22 H0: (p1 – p2) = 0

H1: (p1 – p2) ≠ 0
389
A13.23
A13.24
A13.25
390
A13.26 H0: (p1 – p2) = 0
H1: (p1 – p2) > 0
z = 2.26, p-value = .0119. There is sufficient evidence to infer that the proportion of Americans
earning at least $75,000 is greater in 2008 than in 2006.
A13.27
Two-tail F test: F = 1.26, p-value = .0101 Use Unequal-variances t-test
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
t = -2.54, p-value = .0111. There is enough evidence to conclude that Democrats and Republicans
differ in their intention to vote.
391
A13.28
A13.29 H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
Success = 1
z = -1.08, p-value = .1400. There is not enough evidence to conclude that Democrats and
Republicans differ in how much they thought about the upcoming election for president.
A13.30
LCL = .0470, UCL = .0694
392
Chapter 14
5(10) + 5(15) + 5(20)
14.1a x = = 15
5+5+5
2 2 2 2
SST = ∑ n (x − x) = 5(10 – 15) + 5(15 – 15) + 5(20 – 15) = 250
j j
2
SSE = ∑ (n − 1)s = (5 –1)(50) + (5 – 1)(50) + (5 – 1)(50) = 600
j j
ANOVA Table
Source Degrees of Freedom Sum of Squares Mean Squares F
Treatments k −1 = 2 SST = 250 MST =125 F = 2.50
Error n − k = 12 SSE = 600 MSE = 50
__________________________________________
Total n − k = 14 SS(Total) = 850
10(10) + 10(15) + 10(20)

bx= = 15
5+5+5
2 2 2 2
SST = ∑ n (x − x) = 10(10 – 15) + 10(15 – 15) + 10(20 – 15) = 500
j j
2
SSE = ∑ (n − 1)s = (10 –1)(50) + (10 – 1)(50) + (10 – 1)(50) = 1350
j j
ANOVA Table
Treatments k −1 = 2 SST = 500 MST= 250 F= 5.00
Error n − k = 27 SSE = 1350 MSE = 50
__________________________________________
Total n − k = 29 SS(Total) = 1850
c The F statistic increases.
4(20) + 4(22) + 4(25)

14.2 a x = = 22.33
4+4+4
2 2 2 2
SST = ∑ n (x − x) = 4(20 – 22.33) + 4(22 – 22.33) + 4(25 – 22.33)
j j = 50.67
2
SSE = ∑ (n − 1)s = (4 –1)(10) + (4 – 1)(10) + (4 – 1)(10) = 90
j j
ANOVA Table
Treatments k −1 = 2 SST = 50.67 MST = 25.33 F = 2.53
Error n−k =9 SSE = 90 MSE= 10
b SSE = ∑ (n j − 1)s 2j = (4 –1)(25) + (4 – 1)(25) + (4 – 1)(25) = 225
393
ANOVA Table
Treatments k −1 = 2 SST = 50.67 MST = 25.33 F= 1.01
Error n−k =9 SSE = 225 MSE = 25.0
c The F statistic decreases.
10(30) + 14(35) + 11(33) + 18(40)

14.3 a x = = 35.34
10 + 14 + 11 + 18
SST = ∑ n (x
j j − x ) 2 = 10(30– 35.34) 2 +14(35 – 35.34) 2 + 11(33 –35.34) 2 + 18(40 –35.34) 2 =
737.9
SSE = ∑ (n j − 1)s 2j = (10 –1)(10) + (14 – 1)(10) + (11 – 1)(10) + (18−1)(10) = 490.0
ANOVA Table
Error n − k = 49 SSE = 490.0 MSE = 10.00
10(130) + 14(135) + 11(133) + 18(140)
b x= = 135.34
10 + 14 + 11 + 18
SST = ∑ n (x
j j − x) 2
= 10(130– 135.34) 2 +14(135 – 135.34) 2 + 11(133 –135.34) 2 + 18(140 –135.34) 2 =

737.9
SSE = ∑ (n j − 1)s 2j = (10 –1)(10) + (14 – 1)(10) + (11 – 1)(10) + (18−1)(10) = 610.0
ANOVA Table
Error n − k = 49 SSE = 490.0 MSE = 10.0
c No change
14.4 H 0 : µ1 = µ 2 = µ 3
H1 : At least two means differ.
Rejection region: F > Fα,k −1,n − k = F.05, 2,9 = 4.26
Finance Marketing Management

Mean 2.25 3.25 5.75
Variance 2.25 2.92 2.92
Grand mean = 3.75
394
2 2 2 2
SST = ∑ n (x − x) = 4(2.25–3.75) + 4(3.25 – 3.75) + 4(5.75 –3.75)
j j = 26.00
2
SSE = ∑ (n − 1)s = (4 –1)(2.25) + (4 – 1)(2.92) + (4 – 1)(2.92) = 24.25
j j
ANOVA table
Source Degrees of Freedom Sum of Squares Mean Squares F .
Error n−k =9 SSE = 24.25 MSE = 2.69
F = 4.82, p-value = .0377. There is enough evidence to conclude that there are differences in the
number of job offers between the three MBA majors.
14.5 H 0 : µ1 = µ 2 = µ 3
H 1 : At least two means differ.
Brand 1 Brand 2 Brand 3

Mean 1.33 2.50 2.67
Variance 1.87 2.30 1.47
Grand mean = 2.17
2 2 2 2
SST = ∑ n (x − x) = 6(1.33 – 2.17) + 6(2.50 – 2.17) + 6(2.67 – 2.17)
j j = 6.33
2
SSE = ∑ (n − 1)s = (6 –1)(1.87) + (6 – 1)(2.30) + (6 – 1)(1.47) = 28.17
j j
ANOVA table
Error n − k = 15 SSE = 28.17 MSE = 1.88
F = 1.69, p-value = .2185. There is not enough evidence to conclude that differences exist between
the three brands.
14.6 H 0 : µ1 = µ 2 = µ 3
BA BSc BBA
Mean 3.94 4.78 5.76
Variance 1.26 .92 1.00
Grand mean = 4.83
SST = ∑ n (x
j j − x ) 2 = 5(3.94 – 4.83) 2 + 5(4.78 – 4.83) 2 + 5(5.76 – 4.83) 2 = 8.30
395
SSE = ∑ (n j − 1)s 2j = (5 –1)(1.26) + (5 – 1)(.92) + (5 – 1)(1.00) = 12.73
ANOVA table
Error n − k = 12 SSE = 12.73 MSE = 1.06
F = 3.91, p-value = .0493. There is enough evidence to conclude that students in different degree
program differ in their summer earnings.
14.7 H 0 : µ1 = µ 2 = µ 3
Professors Administrators Students

Mean 6.4 10.4 10.8
Variance 48.3 16.3 37.7
Grand mean = 9.2
2 2 2 2
SST = ∑ n (x − x) = 5(6.4 – 9.2) + 5(10.4 – 9.2) + 5(10.8 – 9.2) = 59.2
j j
2
SSE = ∑ (n − 1)s = (5 –1)(48.3) + (5 – 1)(16.3) + (5 – 1)(37.7) = 409.2
j j
ANOVA table
Treatments k −1 = 2 SST = 59.2 MST = 29.6 F = .87
Error n − k = 12 SSE = 409.2 MSE = 34.1
F = .87, p-value = .4445. There is not enough evidence to conclude that the differing university
communities differ in the amount of spam they receive in their emails.
14.8 H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα,k −1,n − k = F.05,3,8 = 4.07
IBM Dell HP Other

Mean 13.33 11.00 9.67 17.00
Variance 12.33 79.00 22.33 39.00
Grand mean = 12.75
SST = ∑ n (x
j j − x ) 2 = 3(13.33 – 12.75) 2 + 3(11.00 – 12.75) 2 + 3(9.67 – 12.75) 2 + 3(17.00 –
12.75) 2 = 92.92
SSE = ∑ (n j − 1)s 2j = (3 –1)(12.33) + (3 – 1)(79.00) + (3 – 1)(22.33) + (3 – 1)(39.00) = 305.33
396
ANOVA table
Error n−k =8 SSE = 305.33 MSE = 38.17
F = .81, p-value = .5224. There is not enough evidence to conclude that there are differences in
age between the computer brands.
14.9a H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k = F.05,3,77 ≈ 2.76
Grand mean = 65.30
SST = ∑ n (x
j j − x) 2
= 20(68.83 – 65.30) 2 + 26(65.08 – 65.30) 2 + 16(62.01 – 65.30) 2 + 19(64.64 – 65.30) 2

= 430.95
SSE = ∑ (n j − 1)s 2j = (20 –1)(52.28) + (26 – 1)(37.38) + (16 – 1)(63.46) + (19 – 1)(56.88)
= 3903.57
ANOVA table
Error n − k = 77 SSE = 3903.57 MSE = 50.70
F = 2.83, p-value = .0437. There is enough evidence to infer that there are differences in grading
standards between the four high schools.
b The grades of the students at each high school are required to be normally distributed with the
same variance.
c The histograms are approximately bell-shaped and the sample variances are similar.
14.10a H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k = F.05,3,116 ≈ 2.68
Grand mean = 101.0
SST = ∑ n (x
j j − x) 2
= 30(90.17 – 101.0) 2 + 30(95.77 – 101.0) 2 + 30(106.8 – 101.0) 2 + 30(111.17 – 101.0) 2 = 8,464
SSE = ∑ (n j − 1)s 2j
= (30 –1)(991.52) + (30 – 1)(900.87) + (30 – 1)(928.70) + (30 – 1)(1,023.04) = 111,480
397
ANOVA table
Treatments k −1 = 3 SST = 8,464 MST = 2821 F = 2.94
Error n − k = 116 SSE = 111,480 MSE = 961
F = 2.94, p-value = .0363. There is enough evidence to infer that there are differences between the
completion times of the four income tax forms.
b The times for each form must be normally distributed with the same variance.
c The histograms are approximately bell-shaped and the sample variances are similar.
14.11 H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k = F.05,3, 275 ≈ 2.61
Grand mean = 218.0
SST = ∑ n (x
j j − x ) 2 = 41(196.83 – 218.0) 2 + 73(207.78 – 218.0) 2 + 86(223.38 – 218.0) 2
+ 79(232.67 – 218.0) 2 = 45,496
SSE = ∑ (n j − 1)s 2j = (41 –1)(914.05) + (73 – 1)(861.12) + (86 – 1)(1,195.44)
+ (79 – 1)(1,079.81) = 284,400

ANOVA table
Treatments k −1 = 3 SST = 45,496 MST = 151,165 F = 14.66
Error n − k = 275 SSE = 284,400 MSE = 1,034
F = 14.66, p-value = 0. There is enough evidence to infer that there are differences in test scores
between children whose parents have different educational levels.
14.12 H 0 : µ1 = µ 2 = µ 3 = µ 4 = µ 5
Rejection region: F > Fα, k −1, n − k = F.01, 4,120 = 3.48
Grand mean = 173.3
SST = ∑ n (x
j j − x ) 2 = 25(164.6 – 173.3) 2 + 25(185.6 – 173.3) 2 + 25(154.8 – 173.3) 2
+ 25(182.6 – 173.3) 2 + 25(178.9 – 173.3) 2 = 17,251
SSE = ∑ (n j − 1)s 2j = (25 –1)(1,164) + (25 – 1)(1,720) + (25 – 1)(1,114) + (25 – 1)(1,658)
+ (25 – 1)(841.8) = 155,941
398
ANOVA table
Treatments k −1 = 4 SST = 17,251 MST = 4,312.6 F = 3.32
Error n − k = 120 SSE = 155,941 MSE = 1,299.5
F = 3.32, p-value = .0129. There is not enough evidence to allow the manufacturer to conclude
that differences exist between the five lacquers.
b The times until first sign of corrosion for each lacquer must be normally distributed with a
common variance.
c The histograms are approximately bell-shaped with similar sample variances.
14.13 H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k = F.05,3, 297 ≈ 2.61
Grand mean = 16.11
SST = ∑ n (x
j j − x ) 2 = 39(22.21 – 16.11) 2 + 114(18.46 – 16.11) 2 + 81(15.49 – 16.11) 2
+ 67(9.31 – 16.11) 2 = 5,202
SSE = ∑ (n j − 1)s 2j = (39 –1)(121.64) + (114 – 1)(90.39) + (81 – 1)(85.25) + (67 – 1)(65.40) =
25,973
ANOVA table
Error n − k = 297 SSE = 25,973 MSE = 87.45
F = 19.83, p-value = 0. There is enough evidence to infer that there are differences exist between
the four groups.
14.14 H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα, k −1, n − k = F.05, 2,57 ≈ 3.15
Grand mean = 562.6
SST = ∑ n (x
j j − x ) 2 = 20(551.50 – 562.6) 2 + 20(576.75 – 562.6) 2 + 20(559.45 – 562.6) 2 =
6,667
SSE = ∑ (n j − 1)s 2j = (20 –1)(2,741.95) + (20 – 1)(2,641.14) + (20 – 1)(3,129.31) = 161,736
399
ANOVA table
Error n − k = 57 SSE = 161,736 MSE = 2837
F = 1.17, p-value = .3162. There is not enough evidence of a difference between fertilizers in
terms of crop yields.
14.15 H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα, k −1, n − k = F.05, 2, 297 ≈ 3.07
Grand mean = 5.48

2 2 2 2
SST = ∑ n (x − x) = 100(5.81 – 5.48) + 100(5.30 – 5.48) + 100(5.33 – 5.48)
j j = 16.38
2
SSE = ∑ (n − 1)s = (100 –1)(6.22) + (100 – 1)(4.05) + (100 – 1)(3.90) = 1,402.5
j j
ANOVA table
Error n − k = 297 SSE = 1,402.5 MSE = 4.72
F = 1.73, p-value = .1783. There is not enough evidence of a difference between the three
departments.
14.16 H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα, k −1, n − k = F.05,3,116 ≈ 2.68
Grand mean = 77.39
SST = ∑ n (xj j − x ) 2 = 30(74.10 – 77.39) 2 + 30(75.67 – 77.39) 2 + 30(78.50 – 77.39) 2
+ 30(81.30 – 77.39) 2 = 909.42
SSE = ∑ (n j − 1)s 2j = (30 –1)(249.96) + (30 – 1)(184.23) + (30 – 1)(233.36) + (30 – 1)(242.91) =
26,403
ANOVA table
Error n − k = 116 SSE = 26,403 MSE = 227.6
F = 1.33, p-value = .2675. There is not enough evidence of a difference between the four groups of
companies.
400
14.17 H 0 : µ1 = µ 2 = µ 3
a Leaf size: Rejection region: F > Fα,k −1,n − k = F.05, 2,147 ≈ 3.06
Grand mean = 21.49

2 2 2 2
SST = ∑ n (x − x) = 50(24.97 – 21.49) + 50(21.65 – 21.49) + 50(17.84 – 21.49)
j j = 1,270
2
SSE = ∑ (n − 1)s = (50 –1)(48.23) + (50 – 1)(54.54) + (50 – 1)(33.85) = 6,695
j j
ANOVA table
Treatments k −1 = 2 SST = 1,270 MST = 635.0 F = 13.95
Error n − k = 147 SSE = 6,695 MSE = 45.54
F = 13.95, p-value = 0. There is sufficient evidence to conclude that the leaf sizes differ between
the 3 groups.
b Nicotine: Rejection region: F > Fα,k −1,n − k = F.05, 2,147 ≈ 3.06
Grand mean = 13.00

2 2 2 2
SST = ∑ n (x − x) = 50(15.52 – 13.00) + 50(13.39 – 13.00) + 50(10.08 – 13.00)
j j = 753.2
2
SSE = ∑ (n − 1)s = (50 –1)(3.72) + (50 – 1)(3.59) + (50 – 1)(3.83) = 545.54
j j
ANOVA table
Error n − k = 147 SSE = 545.54 MSE = 3.71
F = 101.47, p-value = 0. There is sufficient evidence to infer that the amounts of nicotine differ
between the 3 groups.
14.18 H 0 : µ1 = µ 2 = µ 3 = µ 4
a Ages: Rejection region: F > Fα,k −1,n − k = F.05,3, 291 ≈ 2.61
Grand mean = 36.23
SST = ∑ n (x
j j − x ) 2 = 63(31.30 – 36.23) 2 + 81(34.42 – 36.23) 2 + 40(37.38 – 36.23) 2
+ 111(39.93 – 36.23) 2 = 3,366
SSE = ∑ (n j − 1)s 2j = (63 –1)(28.34) + (81 – 1)(23.20) + (40 – 1)(31.16) + (111 – 1)(72.03) =
12,752
401
ANOVA table
Treatments k −1 = 3 SST =3,366 MST = 1122 F = 25.60
Error n − k = 291 SSE = 12,752 MSE = 43.82
F = 25.60, p-value = 0. There is sufficient evidence to infer that the ages of the four groups of
cereal buyers differ.
b Incomes: Rejection region: F > Fα,k −1,n − k = F.05,3, 291 ≈ 2.61
Grand mean = 39.97
SST = ∑ n (x
j j − x ) 2 = 63(37.22 – 39.97) 2 + 81(38.91 – 39.97) 2 + 40(41.48 – 39.97) 2
+ 111(41.75 – 39.97 ) 2 = 1,008
SSE = ∑ (n j − 1)s 2j = (63 –1)(39.82) + (81 – 1)(40.85) + (40 – 1)(61.38) + (111 – 1)(46.59)
= 13,256
ANOVA table
Treatments k −1 = 3 SST =1,008 MST = 336.0 F = 7.37
Error n − k = 291 SSE = 13,256 MSE = 45.55
F = 7.37, p-value = .0001. There is sufficient evidence to conclude that incomes differ between the
four groups of cereal buyers.
c Education: Rejection region: F > Fα,k −1,n − k = F.05,3, 291 ≈ 2.61
Grand mean = 11.98
SST = ∑ n (x
j j − x ) 2 = 63(11.75 – 11.98) 2 + 81(12.41 – 11.98) 2 + 40(11.73 – 11.98) 2
+ 111(11.89 – 11.98) 2 = 21.71
SSE = ∑ (n j − 1)s 2j = (63 –1)(3.93) + (81 – 1)(3.39) + (40 – 1)(4.26) + (111 – 1)(4.30) = 1,154
ANOVA table
Treatments k −1 = 3 SST =21.71 MST = 7.24 F = 1.82
Error n − k = 291 SSE = 1,154 MSE = 3.97
F = 1.82, p-value = .1428. There is not enough evidence to infer that education differs between the
four groups of cereal buyers.
d Using the F-tests and the descriptive statistics we see that the mean ages and mean household
incomes are in ascending order. For example, Sugar Smacks buyers are younger and earn less than
the buyers of the other three cereals. Cheerio purchasers are older and earn the most.
402
14.19 H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα, k −1, n − k = F.05, 2,57 ≈ 3.15
Grand mean = 146.1
SST = ∑ n (xj j − x ) 2 = 20(153.60 – 146.1) 2 + 20(151.50 – 146.1) 2 + 20(133.25 – 146.1) 2
= 5,011
SSE = ∑ (n j − 1)s 2j = (20 –1)(654.25) + (20 – 1)(924.05) + (20 – 1)(626.83) = 41,898
ANOVA table
Error n − k = 57 SSE = 41,898 MSE = 735.0
F = 3.41, p-value = .0400. There is enough evidence to infer that sales will vary according to price.
14.20 H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα, k −1, n − k = F.05, 2, 232 ≈ 3.07
Grand mean = 19.50

2 2 2 2
SST = ∑ n (x − x) = 61(18.54 – 19.50) + 83(19.34 – 19.50) + 91(20.29 – 19.50)
j j = 114.5
2
SSE = ∑ (n − 1)s = (61 –1)(177.95) + (83 – 1)(171.42) + (91 – 1)(297.50) = 51,508
j j
ANOVA table
Error n − k = 232 SSE = 51,508 MSE = 222.0
F = .26, p-value = .7730. There is not enough evidence of a difference between the three segments.
14.21 H0: µ1 = µ2 = µ3 =µ4

H1: At least two means differ.
403
F = 10.54, p-value 6.42E-07 ≈ 0. There is enough evidence to infer that the number of years of
education differ in the four years.
14.22 H 0: µ 1 = µ 2 = µ 3
F = 31.86, p-value ≈ 3.06E-140. There is enough evidence to infer that the amount of television
differs by race.
14.23 H 0: µ 1 = µ 2 = µ 3 = µ 4 = µ 5
404
F = 63.15, p-value ≈ 0. There is enough evidence to infer that income differs by completed degree.
14.24 H 0: µ 1 = µ 2 = µ 3 = µ 4
F = .33, p-value = .8005. There is not enough evidence to infer that the amount of time spent at
work differs over the four years.
14.25 H 0: µ 1 = µ 2 = µ 3 = µ 4 = µ 5
405
F = 13.18, p-value ≈ 0. There is enough evidence to infer that income differs by marital status.
14.26 H 0: µ 1 = µ 2 = µ 3 = µ 4
F = .496, p-value = .6852. There is not enough evidence to infer that the amount of television
differs by year.
14.27 H 0: µ 1 = µ 2 = µ 3 = µ 4 = µ 5 = µ 6 = µ 7
406
F = 3.60, p-value = .0015. There is enough evidence to infer that income differs between the seven
political views.
14.28 H 0: µ 1 = µ 2 = µ 3
F = 11.59, p-value ≈ 0. There is enough evidence to infer that income differs between the three
political affiliations.
407
14.29 H 0: µ 1 = µ 2 = µ 3
F = 1.58, p-value = .2062. There is not enough evidence to infer that the number of days reading a
newspaper differs between the three political affiliations.
14.30 H0: µ1 = µ2 = µ3 = µ4 = µ5 =µ6 = µ7 = µ8

F = 17.10, p-value ≈ 0. There is enough evidence to infer that income differs between the
completed degrees.
408
14.31 H 0: µ 1 = µ 2 = µ 3 = µ 4 = µ 5 = µ 6
F = 13.94, p-value ≈ 0. There is enough evidence to infer that the amount of education differs by
marital status.
14.32 H 0: µ 1 = µ 2 = µ 3
F = 37.47, p-value ≈ 0. There is enough evidence to conclude that intention to vote differs by
political affiliation.
409
14.33 a α = .05: t α / 2, n − k = t.025,27 = 2.052
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.052 700⎜ + ⎟ = 24.28
⎜ n i n j ⎟ ⎝ 10 10 ⎠
⎝ ⎠
Treatment Means Difference
i = 1, j = 2 128.7 101.4 27.3
i = 1, j = 3 128.7 133.7 −5.0
i = 2, j = 3 101.4 133.7 −32.3
Conclusion: µ 2 differs from µ1 and µ 3 .
b C = 3(2)/2 = 3, α E = .05, α = α E / C = .0167: t α / 2,n − k = t .0083, 27 = 2.552 (from Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.552 700⎜ + ⎟ = 30.20
⎜ n i n j ⎟ ⎝ 10 10 ⎠
⎝ ⎠
i = 1, j = 2 128.7 101.4 27.3
i = 1, j = 3 128.7 133.7 −5.0
i = 2, j = 3 101.4 133.7 −32.3
Conclusion: µ 2 and µ 3 differ.
MSE 700
c q α (k, ν) = q .05 (3,27) ≈ 3.53 ϖ = q α (k, ν) = 3.53 = 29.53
ng 10

i = 1, j = 2 128.7 101.4 27.3
i = 1, j = 3 128.7 133.7 −5.0
i = 2, j = 3 101.4 133.7 −32.3
Conclusion: µ 2 and µ 3 differ.
14.34 a α = .05: t α / 2, n − k = t.025, 20 = 2.086
⎛ 1 ⎞
LSD = t α / 2, n − k MSE⎜ +
1 ⎟ = 2.086 125⎛⎜ 1 + 1 ⎞⎟ = 14.75
⎜ n i n j ⎟ ⎝ 5 5 ⎠
⎝ ⎠
i = 1, j = 2 227 205 22
i = 1, j = 3 227 219 8
i = 1, j = 4 227 248 −21
i = 1, j = 5 227 202 25
i = 2, j = 3 205 219 −14
i = 2, j = 4 205 248 −43
410
i = 2, j = 5 205 202 3
i = 3, j = 4 219 248 −29
i = 3, j = 5 219 202 17
i = 4, j = 5 248 202 46
Conclusion: The following pairs of means differ. µ1 and µ 2 , µ1 and µ 4 , µ1 and µ 5 , µ 2 and
µ 4 , µ 3 and µ 4 , µ 3 and µ 5 , and µ 4 and µ 5 .
b C = 5(4)/2 = 10, α E = .05, α = α E / C = .005 t α / 2,n − k = t .0025 , 20 = 3.153 (from Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 3.153 125⎜ + ⎟ = 22.30
⎜ n i n j ⎟ ⎝ 5 5 ⎠
⎝ ⎠
i = 1, j = 2 227 205 22
i = 1, j = 3 227 219 8
i = 1, j = 4 227 248 −21
i = 1, j = 5 227 202 25
i = 2, j = 3 205 219 −14
i = 2, j = 4 205 248 −43
i = 2, j = 5 205 202 3
i = 3, j = 4 219 248 −29
i = 3, j = 5 219 202 17
i = 4, j = 5 248 202 46
Conclusion: The following pairs of means differ. µ1 and µ 5 , µ 2 and µ 4 , µ 3 and µ 4 , and µ 4 and
µ5 .
MSE 125
c q α (k, ν) = q .05 (5, 20) = 4.23 ϖ = q α (k, ν) = 4.23 = 21.15
ng 5
i = 1, j = 2 227 205 22
i = 1, j = 3 227 219 8
i = 1, j = 4 227 248 −21
i = 1, j = 5 227 202 25
i = 2, j = 3 205 219 −14
i = 2, j = 4 205 248 −43
i = 2, j = 5 205 202 3
i = 3, j = 4 219 248 −29
i = 3, j = 5 219 202 17
i = 4, j = 5 248 202 46
411
Conclusion: The following pairs of means differ. µ1 and µ 2 , µ1 and µ 5 , µ 2 and µ 4 , µ 3 and µ 4 ,
and µ 4 and µ 5 .
MSE 1.88
14.35 q α (k, ν) = q .05 (3, 15) = 3.67 ϖ = q α (k, ν) = 3.67 = 2.05
ng 6

i = 1, j = 2 1.33 2.50 −1.17
i = 1, j = 3 1.33 2.67 −1.34
i = 2, j = 3 2.50 2.67 −.17
There are no differences.
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
14.36 a LSD = t α / 2, n − k MSE⎜ + ⎟ = 1.782 1.06⎜ + ⎟ = 1.16
⎜ n i n j ⎟ ⎝ 5 5 ⎠
⎝ ⎠
i = 1, j = 2 3.94 4.78 −.84
i = 1, j = 3 3.94 5.76 −1.82
i = 2, j = 3 4.78 5.76 −1.02
Means of BAs and BBAs differ.
b C = 3(2)/2 = 3, α E = .10, α = α E / C = .0333: t α / 2,n − k = t .0167 ,12 = 2.404 (from Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.404 1.06⎜ + ⎟ = 1.57
⎜ n i n j ⎟ ⎝ 5 5 ⎠
⎝ ⎠
i = 1, j = 2 3.94 4.78 −.84
i = 1, j = 3 3.94 5.76 −1.82
i = 2, j = 3 4.78 5.76 −1.02
Means of BAs and BBAs differ.
14.37 LSD method: C = 4(3)/2 = 6, α E = .05, α = α E / C = .0083
⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ (LSD must be calculated for each pair of treatments.)
⎜ n i n j ⎟
⎝ ⎠
k 4
Tukey’s method: n g = = = 19.6 (rounded to 20)
1 1 1 1 1 1 1
+ + ... + + + +
n1 n 2 nk 20 26 16 19
50.70
q α (k, ν) = q .05 (4, 77) ≈ 3.74 ϖ = 3.74 = 5.95
20
412
Treatment Means Difference LSD
i = 1, j = 2 68.83 65.08 3.75 5.74
i = 1, j = 3 68.83 62.01 6.82 6.47
i = 1, j = 4 68.83 64.64 4.19 6.18
i = 2, j = 3 65.08 62.01 3.07 6.13
i = 2, j = 4 65.08 64.64 .44 5.82
i = 3, j = 4 62.01 64.64 −2.63 6.55
a The mean grades from high schools A and C differ.
b The mean grades from high schools A and C differ.
961.0
14.38 Tukey’s method: q α (k, ν) = q .05 (4, 116) ≈ 3.68 ϖ = 3.68 = 20.83
30
LSD method with the Bonferroni adjustment: C = 4(3)/2 = 6, α E = .05, α = α E / C = .0083
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.64 961.0⎜ + ⎟ = 21.13
⎜ n i n j ⎟ ⎝ 30 30 ⎠
⎝ ⎠
Treatment Means Difference Tukey LSD
i = 1, j = 2 90.17 95.77 −5.60 20.83 21.13
i = 1, j = 3 90.17 106.8 −16.67 20.83 21.13
i = 1, j = 4 90.17 111.17 −21.00 20.83 21.13
i = 2, j = 3 95.77 106.8 −11.07 20.83 21.13
i = 2, j = 4 95.77 111.17 −15.40 20.83 21.13
i = 3, j = 4 106.8 111.17 −4.33 20.83 21.13
a The means for Forms 1 and 4 differ.
b No means differ.
14.39a H 0 : µ1 = µ 2 = µ 3 = µ 4
10(61.60) + 10(57.30) + 10(61.80) + 10(51.80)
x= = 58.13
10 + 10 + 10 + 10
SST = ∑ n (xj j − x ) 2 = 10(61.6–58.13) 2 +10(57.3–58.13) 2 + 10(61.8–58.13) 2 + 10(51.8–
58.13) 2 = 662.7
SSE = ∑ (n j − 1)s 2j = (10 –1)(80.49) + (10 – 1)(70.46) + (10 – 1)(22.18) + (10−1)(75.29) = 2,236
413
ANOVA Table
Treatments k −1 = 3 SST = 662.7 MST= 220.9 F = 3.56
Error n − k = 36 SSE = 2,236 MSE = 62.11
F = 3.56, p-value = .0236. There is enough evidence to infer that differences exist between the
flares with respect to burning times.
LSD method: C = 4(3)/2 = 6, α E = .05, α = α E / C = .0083 t α / 2,n − k = t .0042 ,36 = 2.794 (from
Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.794 62.11⎜ + ⎟ = 9.85
⎜ n i n j ⎟ ⎝ 10 10 ⎠
⎝ ⎠
62.11
Tukey’s method: q α (k, ν) = q .05 (4, 136) ≈ 3.79 ϖ = 3.79 = 9.45
10
i = 1, j = 2 61.6 57.3 4.3
i = 1, j = 3 61.6 61.8 −.2
i = 1, j = 4 61.6 51.8 9.8
i = 2, j = 3 57.3 61.8 −4.5
i = 2, j = 4 57.3 51.8 5.5
i = 3, j = 4 61.8 51.8 10.0
b. The means of flares C and D differ.
c The means of flares A and D, and C and D differ.
14.40 a LSD method: C = 5(4)/2 = 10, α E = .05, α = α E / C = .005
t α / 2,n − k = t .0025 ,120 = 2.860 (from Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.860 1300⎜ + ⎟ = 29.17
⎜ n i n j ⎟ ⎝ 25 25 ⎠
⎝ ⎠
1300
b Tukey’s method: q α (k, ν) = q .05 (5,120) = 3.92 ϖ = 3.92 = 28.27
25
i = 1, j = 2 164.6 185.6 −21.0
i = 1, j = 3 164.6 154.8 9.8
i = 1, j = 4 164.6 182.6 −18.0
i = 1, j = 5 164.6 178.9 −14.3
i = 2, j = 3 185.6 154.8 30.8
i = 2, j = 4 185.6 182.6 3.0
i = 2, j = 5 185.6 178.9 6.7
414
i = 3, j = 4 154.8 182.6 −27.8
i = 3, j = 5 154.8 178.9 −24.1
i = 4, j = 5 182.6 178.9 3.7
a The means of lacquers 2 and 3 differ
b The means of lacquers 2 and 3 differ.
14.41a H 0 : µ1 = µ 2 = µ 3
30(53.17) + 30(49.37) + 30(44.33)
x= = 48.96
30 + 30 + 30
2 2 2 2
SST =∑ n (x − x) = 30(53.17–48.96) +30(49.37–48.96) + 30(44.33–48.96)
j j = 1,178
2
SSE = ∑ (n − 1)s = (30 –1)(194.6) + (30 – 1)(152.6) + (30 – 1)(129.9) = 13,836
j j
ANOVA Table
Treatments k −1 = 2 SST = 1,178 MST = 589.0 F = 3.70
Error n − k = 87 SSE = 13,836 MSE = 159.0
F = 3.70, p-value = .0286. There is enough evidence to infer that speed of promotion varies
between the three sizes of engineering firms.
159.0
b q α (k, ν) = q .05 (3,87) ≈ 3.40 ϖ = 3.40 = 7.83
30
i = 1, j = 2 53.17 49.37 3.80
i = 1, j = 3 53.17 44.44 8.83
i = 2, j = 3 49.37 44.33 5.04
The means of small and large firms differ. Answer (v) is correct.
2,838
14.42 Tukey’s method: q α (k, ν) = q .05 (3,57) ≈ 3.40 ϖ = 3.40 = 40.50
20
LSD method: C = 3(2)/2 = 3, α E = .05, α = α E / C = .0167 t α / 2,n − k = t .0083 ,57 = 2.466 (from
Excel)
⎛ 1 1 ⎞ ⎛ 1 1 ⎞
LSD = t α / 2, n − k MSE⎜ + ⎟ = 2.466 2,838⎜ + ⎟ = 41.54
⎜ n i n j ⎟ ⎝ 20 20 ⎠
⎝ ⎠
415
i = 1, j = 2 551.5 576.8 −25.3
i = 1, j = 3 551.5 559.5 −8.0
i = 2, j = 3 576.8 559.5 17.3
a There are no differences.
b There are no differences.
14.43
Using Tukey’s method or the Bonferroni adjustment the following pairs of years differ 2002 and
2004, 2004 and 2008.
14.44
Blacks differ from Whites and from Other.
416
14.45
Using Tukey’s method or the Bonferroni adjustment the only pair that don’t differ are Degress 1
and 2.
14.46
Using Tukey’s method the following pairs of means differ: Married and Separated, Married and
Never married, and Divorced and Separated
417
14.47
The only pair of means that differ are Extremely liberal and Slightly conservative.
14.48
Using Tukey’s method or the Bonferroni adjustment the following pairs of means differ:
Democrat and Republican and Republican and Independent.
418
14.49
The following pairs of means differ: Married and Divorced, Married and Separated, Widowed and
Separated, Separated and Never married, and Separated and Partnered.
14.50
Using Tukey’s method all three pairs of means differ. Using the Bonferroni adjustment Democrats
and Independents and Republicans and Independents differ.
14.51 ANOVA Table

Source Degrees of Freedom Sum of Squares Mean Squares F .
Treatments 2 100 50.00 24.04
Blocks 6 50 8.33 4.00
Error 12 25 2.08
Total 20 175
a Rejection region: F > Fα,k −1,n − k − b +1 = F.05, 2,12 = 3.89
419
Conclusion: F = 24.04, p-value = .0001. There is enough evidence to conclude that the treatment
means differ.
b Rejection region: F > Fα,b −1, n − k − b +1 = F.05,6,12 = 3.00
F = 4.00, p-value = .0197. There is enough evidence to conclude that the block means differ.
14.52 ANOVA Table

Treatments 4 1,500 375.0 16.50
Blocks 11 1,000 90.91 4.00
Error 44 1,000 22.73
Total 59 3,500
a Rejection region: F > Fα,k −1,n − k − b +1 = F.01, 4,44 ≈ 3.83
F = 16.50, p-value = 0. There is enough evidence to conclude that the treatment means differ.
b Rejection region: F > Fα,b−1,n −k −b+1 = F.01,11, 44 ≈ 2.80
Conclusion: F = 4.00, p-value = .0005. There is enough evidence to conclude that the block
means differ.
14.53 ANOVA Table

Treatments 3 275 91.67 7.99
Blocks 9 625 69.44 6.05
Error 27 310 11.48
Total 39 1,210
a Rejection region: F > Fα,k −1,n − k − b +1 = F.01,3, 27 = 4.60
means differ.
b Rejection region: F > Fα,b −1, n − k − b +1 = F.01,9, 27 = 3.15
Conclusion: F = 6.05, p-value = .0001. There is enough evidence to conclude that the block
means differ.
14.54 Rejection region: F > Fα,k −1,n − k − b +1 = F.05, 2,14 = 3.74
420
a ANOVA Table
Treatments 2 1,500 750.0 7.00
Blocks 7 500 71.43 .67
Error 14 1,500 107.1
Total 23 3,500
means differ.
b ANOVA Table
Treatments 2 1,500 750.0 10.50
Blocks 7 1,000 142.86 2.00
Error 14 1,000 71.43
Total 23 3,500
Conclusion: F = 10.50, .0016. There is enough evidence to conclude that the treatment means
differ
c ANOVA Table
Treatments 2 1,500 750 21.00
Blocks 7 1,500 214.3 6.00
Error 14 500 35.71
Total 23 3,500
means differ
d The test statistic increases.
14.55 a k = 3, b = 5, Grand mean = 10.4

SS(Total) =
k b
∑ ∑ (x ij − x ) 2 = (7 − 10.4) 2 + (10 − 10.4) 2 + (12 − 10.4) 2 + (9 − 10.4) 2 + (12 − 10.4) 2

j=1 i =1
+ (12 −10.4) 2 + (8 −10.4) 2 + (16 −10.4) 2 + (13 −10.4) 2 + (10 −10.4) 2
+ (8 −10.4) 2 + (9 −10.4) 2 + (13 −10.4) 2 + (6 −10.4) 2 + (11 −10.4) 2 = 99.6

k
SST = ∑ b(x[T] j − x ) 2 = 5[(10 − 10.4) 2 + (11.8 − 10.4) 2 + (9.4 − 10.4) 2 ] = 15.6
j=1
421
SSB =
∑ k(x[B] i − x ) 2 = 3[(9 − 10.4) 2 + (9 − 10.4) 2 + (13.7 − 10.4) 2 + (9.3 − 10.4) 2 + (11 − 10.4) 2 ] = 48.3
i =1
SSE = SS(Total) – SST – SSB = 99.6− 15.6 – 48.3 = 35.7

b SS(Total) =
k b
∑ ∑ (x ij − x ) 2 = (7 − 10.4) 2 + (10 − 10.4) 2 + (12 − 10.4) 2 + (9 − 10.4) 2 + (12 − 10.4) 2

j=1 i =1
+ (12 −10.4) 2 + (8 −10.4) 2 + (16 −10.4) 2 + (13 −10.4) 2 + (10 −10.4) 2
+ (8 −10.4) 2 + (9 −10.4) 2 + (13 −10.4) 2 + (6 −10.4) 2 + (11 −10.4) 2 = 99.6

k
SST = ∑ n (x j j − x ) 2 = 5(10 − 10.4) 2 + 5(11.8 − 10.4) 2 + 5(9.4 − 10.4) 2 = 15.6
j=1
SSE = SS(Total) – SST = 99.6 – 15.6 = 84.0

c The variation between all the data is the same for both designs.
d The variation between treatments is the same for both designs.
e Because the randomized block design divides the sum of squares for error in the one-way
analysis of variance into two parts.
14.56 a k = 4, b = 3, Grand mean = 5.6

SS(Total) =
k b
∑ ∑ (x ij − x ) 2 = (6 − 5.6) 2 + (8 − 5.6) 2 + (7 − 5.6) 2 + (5 − 5.6) 2 + (5 − 5.6) 2 + (6 − 5.6) 2

j=1 i =1
+ (4 − 5.6) 2 + (5 − 5.6) 2 + (5 − 5.6) 2 + (4 − 5.6) 2 + (6 − 5.6) 2 + (6 − 5.6) 2 = 14.9

k
SST = ∑ b(x[T] j − x ) 2 = 3[(7 − 5.6) 2 + (5.3 − 5.6) 2 + (4.7 − 5.6) 2 + (5.3 − 5.6) 2 ] = 8.9
j=1
b
SSB = ∑ k(x[B] i − x ) 2 = 4[(4.8 − 5.6) 2 + (6 − 5.6) 2 + (6 − 5.6) 2 = 4.2
i =1
SSE = SS(Total) – SST – SSB = 14.9− 8.9 – 4.2 = 1.8

b SS(Total) =
k b
∑ ∑ (x ij − x ) 2 = (6 − 5.6) 2 + (8 − 5.6) 2 + (7 − 5.6) 2 + (5 − 5.6) 2 + (5 − 5.6) 2 + (6 − 5.6) 2

j=1 i =1
+ (4 − 5.6) 2 + (5 − 5.6) 2 + (5 − 5.6) 2 + (4 − 5.6) 2 + (6 − 5.6) 2 + (6 − 5.6) 2 = 14.9
422
k
SST = ∑ b(x[T] j − x ) 2 = 3[(7 − 5.6) 2 + (5.3 − 5.6) 2 + (4.7 − 5.6) 2 + (5.3 − 5.6) 2 ] = 8.9
j=1
SSE = SS(Total) – SST = 14.9 – 8.9 = 6.0

c The variation between all the data is the same for both designs.
d The variation between treatments is the same for both designs.
e Because the randomized block design divides the sum of squares for error in the one-way
analysis of variance into two parts.
14.57 H0 : µ1 = µ2 = µ 3
Rejection region: F > Fα,k −1,n − k − b +1 = F.05, 2,6 = 5.14
k = 3, b = 4, Grand mean = 2.38

k b
SS(Total) = ∑ ∑ (x ij − x ) 2 = (1.4 − 2.38) 2 + (3.1 − 2.38) 2 + (2.8 − 2.38) 2 + (3.4 − 2.38) 2
j=1 i =1
+ (1.5 − 2.38) 2 + (2.6 − 2.38) 2 + (2.1 − 2.38) 2 + (3.6 − 2.38) 2
+ (1.3 − 2.38) 2 + (2.4 − 2.38) 2 + (1.5 − 2.38) 2 + (2.9 − 2.38) 2 = 7.30

k
SST = ∑ b(x[T] j − x ) 2 = 4[(2.68 − 2.38) 2 + (2.45 − 2.38) 2 + (2.03 − 2.38) 2 ] = .87
j=1
b
SSB = ∑ k(x[B] i − x ) 2 = 3[(1.4 − 2.38) 2 + (2.7 − 2.38) 2 + (2.1 − 2.38) 2 + (3.3 − 2.38) 2 ] = 5.91
i =1
SSE = SS(Total) – SST – SSB = 7.30 − .87 – 5.91 = .52

ANOVA Table
Treatments 2 .87 .44 5.06
Blocks 3 5.91 1.97 22.64
Error 6 .52 .087
Total 11 7.30
F = 5.08, p-value = .0512. There is not enough evidence to conclude that there are differences
between the subjects being measured.
14.58 H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k − b +1 = F.01,3,12 = 5.95
423
k = 4, b = 5, Grand mean = 8.3
k b
SS(Total) = ∑ ∑ (x ij − x)2
j=1 i =1
= (5 − 8.3)2 + (4 − 8.3)2 + (6 − 8.3)2 + (7 − 8.3)2 + (9 − 8.3)2
+ (2 − 8.3) 2 + (7 − 8.3) 2 + (12 − 8.3) 2 + (11 − 8.3) 2 + (8 − 8.3) 2
+ (6 − 8.3) 2 + (8 − 8.3) 2 + (9 − 8.3) 2 + (16 − 8.3) 2 + (15 − 8.3) 2
+ (8 − 8.3) 2 + (10 − 8.3) 2 + (2 − 8.3) 2 + (7 − 8.3) 2 + (14 − 8.3) 2 = 286.2
k
SST = ∑ b(x[T] j − x ) 2 = 5[(6.2 − 8.3) 2 + (8.0 − 8.3) 2 + (10.8 − 8.3) 2 + (8.2 − 8.3) 2 ] = 53.8
j=1
b
2
SSB = ∑= k4([(x5[B.25] −− 8x.)3)2 + (7.25 − 8.3)2 + (7.25 − 8.3)2 + (10.25 − 8.3)2 + (11.5 − 8.3)2 ] = 102.2
i =1
i
SSE = SS(Total) – SST – SSB = 286.2 – 53.8 – 102.2 = 130.2

ANOVA Table
Treatments 3 53.8 17.93 1.65
Blocks 4 102.2 25.55 2.35
Error 12 130.2 10.85
Total 19 286.2
F = 1.65, p-value = .2296. There is not enough evidence to conclude there are differences between
the four diets.
14.59 ANOVA Table

Treatments 2 204.2 102.11 4.54
Blocks 11 1150.2 104.57 4.65
Error 22 495.1 22.51
a H 0 : µ1 = µ 2 = µ 3
F = 4.54, p-value = .0224. There is enough evidence to conclude that there are differences between
the three couriers.
424
b H 0 : µ1 = µ 2 = … = µ12
F = 4.65, p-value = .0011. The block means differ; the practitioner used the correct design.
14.60 ANOVA Table

Treatments 2 7,131 3,566 123.36
Blocks 19 177,465 9,340 323.16
Error 38 1,098 28.90
a H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα,k −1,n − k − b +1 = F.05, 2,38 ≈ 3.23
F = 123.36, p-value = 0. There is sufficient evidence to conclude that the three fertilizers differ
with respect to crop yield.
b F = 323.16, p-value = 0. There is sufficient evidence to indicate that there are differences
between the plots.
14.61 ANOVA Table

Treatments 2 10.26 5.13 .86
Blocks 19 3,020 159.0 26.64
Error 38 226.7 5.97
a H 0 : µ1 = µ 2 = µ 3
Rejection region: F > Fα,k −1,n − k − b +1 = F.05,2,38 ≈ 3.23
F = .86, p-value = .4313. There is not enough evidence to conclude that there are differences in
sales ability between the holders of the three degrees.
b H 0 : µ1 = µ 2 = … = µ 20
F = 26.64, p-value = 0. There is sufficient evidence to indicate that there are differences between
the blocks of students. The independent samples design would not be recommended.
c The commissions for each type of degree are required to be normally distributed with the same
variance.
d The histograms are bell shaped and the sample variances are similar.
425
14.62 ANOVA Table
Treatments 3 4,206 1,402 21.16
Blocks 29 126,843 4,374 66.02
Error 87 5,764 66.25
a H 0 : µ1 = µ 2 = µ 3 = µ 4
Rejection region: F > Fα,k −1,n − k − b +1 = F.01,3,87 ≈ 4.01
F = 21.16, p-value = 0. There is sufficient evidence to conclude that differences in completion

times exist between the four forms.
b H 0 : µ1 = µ 2 = … = µ 30
the taxpayers, which tells us that this experimental design is recommended.
14.63 H 0 : µ1 = µ 2 = µ 3 = µ 4 = µ 5 = µ 6 = µ 7
ANOVA Table
Treatments 6 28,674 4,779 11.91
Blocks 199 209,835 1,054 2.63
Error 1194 479,125 401.3
F = 11.91, p-value = 0. There is enough evidence to conclude that there are differences in time
spent listening to music between the days of the week
14.64 ANOVA Table

Treatments 4 1,406 351.6 10.72
Blocks 35 7,310 208.9 6.36
Error 140 4,594 32.81
a H 0 : µ1 = µ 2 = µ 3 = µ 4 = µ 5
F = 10.72, p-value = 0. There is enough evidence to infer differences between medical specialties.
b H 0 : µ1 = µ 2 = … = µ 36
426
the physicians’ ages, which tells us that this experimental design is recommended.
14.65 H 0 : µ1 = µ 2 = µ 3 = µ 4
ANOVA Table
Treatments 3 563.8 187.9 15.06
Blocks 20 1,327 66.37 5.32
Error 60 748.7 12.48
F = 15.06, p-value = 0. There is enough evidence to infer differences in grading standards between
the four high schools.
14.66 H 0 : µ1 = µ 2 = µ 3 = µ 4
F = 44.74, p-value 3.30E-28 ≈ 0 . There is enough evidence to conclude that the means differ.
14.67 H 0 : µ1 = µ 2 = µ 3 = µ 4
427
F = 23.04, p-value 9.06E-15 ≈ 0 . There is enough evidence to conclude that the means differ.
14.68a ANOVA Table

Treatments 11 12,045 1095 8.23
Error 228 30,405 133
Total 239 42,450
b. Rejection region: F > Fα,k −1,n − k = F.01,11,228 ≈ 2.25
F = 8.53. There is enough evidence to conclude that the treatment means differ.
c. ANOVA Table
Factor A 2 1,560 780 5.86
Factor B 3 2,880 960 7.18
Interaction 6 7,605 1268 9.53
Error 228 30,405 133
Total 239 42,450
Interaction: Rejection region: F > Fα,( a −1) b−1), n − k = F.01,6,228 ≈ 2.80
F = 9.53. There is enough evidence to infer that factors A and B interact.
14.69 ANOVA Table

Factor A 3 203 67.67 .72
Factor B 2 859 429.5 4.60
Interaction 6 513 85.5 .92
Error 84 7845 93.39
428
Total 95 9420
a Rejection region: F > Fα,(a −1)( b −1), n −ab = F.05,6,84 ≈ 2.25
F = .92. There is not enough evidence to conclude that factors A and B interact.
b Rejection region: F > Fα,a −1,n −ab = F.05,3,84 ≈ 2.76
F = .72. There is not enough evidence to conclude that differences exist between the levels of
factor A.
c Rejection region: F > Fα,b −1,n −ab = F.05, 2,84 ≈ 3.15
F = 4.60. There is enough evidence to conclude that differences exist between the levels of factor
B.
14.70 ANOVA Table

A B C D E F G
23 ANOVA
24 Source of Variation SS df MS F P-value F crit
25 Sample 5.33 1 5.33 1.23 0.2995 5.32
26 Columns 56.33 1 56.33 13.00 0.0069 5.32
27 Interaction 1.33 1 1.33 0.31 0.5943 5.32
28 Within 34.67 8 4.33
29
30 Total 97.67 11
a F = .31, p-value = .5943. There is not enough evidence to conclude that factors A and B interact.
b F = 1.23, p-value = .2995. There is not enough evidence to conclude that differences exist
between the levels of factor A.
c F = 13.00, p-value = .0069. There is enough evidence to conclude that differences exist between
the levels of factor B.
4.71 ANOVA Table

A B C D E F G
29 ANOVA
31 Sample 177.25 2 88.63 9.99 0.0012 3.55
32 Columns 0.38 1 0.38 0.04 0.8394 4.41
33 Interaction 9.25 2 4.63 0.52 0.6025 3.55
34 Within 159.75 18 8.88
35
36 Total 346.63 23
a F = .52, p-value = .6025. There is not enough evidence to conclude that factors A and B interact.
b F = 9.99, p-value = .0012. There is enough evidence to conclude that differences exist between
the levels of factor A.
c F = .04, p-value = .8394. There is not enough evidence to conclude that differences exist
between the levels of factor B.
429
14.72 ANOVA Table
A B C D E F G
35 ANOVA
37 Sample 135.85 3 45.28 4.49 0.0060 2.7318
38 Columns 151.25 1 151.25 15.00 0.0002 3.9739
39 Interaction 6.25 3 2.08 0.21 0.8915 2.7318
40 Within 726.20 72 10.09
41
42 Total 1019.55 79
The test for interaction yields (F = .21, p-value = .8915) and the test for the differences between
educational levels (F = 4.49, p-value = .0060) is the same as in Example 14.4. However, in this
exercise there is evidence of a difference between men and women (F = 15.00, p-value = .0002).
14.73 ANOVA Table

A B C D E F G
35 ANOVA
37 Sample 345.85 3 115.28 11.43 3.25E-06 2.7318
38 Columns 61.25 1 61.25 6.07 0.0161 3.9739
39 Interaction 72.25 3 24.08 2.39 0.0760 2.7318
40 Within 726.20 72 10.09
41
42 Total 1205.55 79
Compared to Example 14.4, the test for interaction has the same conclusion, although the value of
F is larger and the p-value is smaller. Moreover, the mean number of jobs differs between the
educational levels (F = 11.43, p-value = 0) and between men and women (F = 6.07, p-value =
.0161).
14.74 a There are 12 treatments.

b There are two factors, tax form and income group.
c There are a = 4 forms and b = 3 income groups.
A B C D E F G
28
29 ANOVA
31 Sample 6719 2 3359.4 4.11 0.0190 3.08
32 Columns 6280 3 2093.3 2.56 0.0586 2.69
33 Interaction 5102 6 850.3 1.04 0.4030 2.18
34 Within 88217 108 816.8
35
36 Total 106317 119
d F = 1.04, p-value = .4030. There is not enough evidence to conclude that forms and income
groups interact
430
e F = 2.56, p-value = .0586. There is not enough evidence to conclude that differences exist
between the forms.
f F = 4.11, p-value = .0190. There is enough evidence to conclude that differences exist between
the three income groups.
14.75 a Detergents and temperatures

b The response variable is the whiteness score.
c There are a = 5 detergents and b = 3 temperatures.
A B C D E F G
29 ANOVA
31 Sample 3937 2 1968.5 17.82 0.0000 3.06
32 Columns 2967 4 741.9 6.72 0.0001 2.44
33 Interaction 2452 8 306.5 2.78 0.0071 2.01
34 Within 14910 135 110.4
35
36 Total 24267 149
d Test for interaction: F = 2.78, p-value = .0071. There is sufficient evidence to conclude that
detergents and temperatures interact. The F-tests in Parts e and f are irrelevant.
14.76 a Factor A is the drug mixture and factor B is the schedule.

b The response variable is the improvement index.
c There are a = 4 drug mixtures and b = 2 schedules.
A B C D E F G
23 ANOVA
25 Sample 14.40 1 14.40 0.57 0.4548 4.15
26 Columns 581.80 3 193.93 7.71 0.0005 2.90
27 Interaction 548.60 3 182.87 7.27 0.0007 2.90
28 Within 804.80 32 25.15
29
30 Total 1949.60 39
d Test for interaction: F = 7.27, p-value = .0007. There is sufficient evidence to conclude that the
schedules and drug mixtures interact. There is sufficient evidence to conclude that detergents and
temperatures interact. The F-tests in Parts e and f are irrelevant.
14.77 a There are 2 factors--class configuration and time period.

b The response variable is the number of times students ask and answer questions.
c There are 2 levels of class configuration and 3 levels of time period.
431
ANOVA
Source of Variation SS df MS F P-value F crit
Sample 13.33 1 13.33 1.58 0.2203 4.26
Columns 46.67 2 23.33 2.77 0.0826 3.40
Interaction 206.67 2 103.33 12.28 0.0002 3.40
Within 202.00 24 8.42
Total 468.67 29
d Interaction: F = 12.28, p-value = .0002. There is sufficient evidence to conclude that the class
configuration and time interact. The other two F-tests are invalid.
14.78
A B C D E F G
23 ANOVA
25 Sample 16.04 1 16.04 14.74 0.0005 4.11
26 Columns 6.77 1 6.77 6.22 0.0173 4.11
27 Interaction 0.025 1 0.025 0.023 0.8814 4.11
28 Within 39.17 36 1.09
29
30 Total 62.00 39
The p-values for interaction, machines, and alloys are .8814, .0173, .0005, and, respectively. Both
machines and alloys are sources of variation.
14.79
A B C D E F G
35 ANOVA
37 Sample 0.000309 3 0.000103 0.66 0.5798 2.90
38 Columns 0.000515 1 0.000515 3.33 0.0775 4.15
39 Interaction 0.000183 3 0.000061 0.39 0.7584 2.90
40 Within 0.004953 32 0.000155
41
42 Total 0.005959 39
The p-values for interaction, devices, and alloys are .7584, .0775, .5798, and, respectively. There
are no sources of variation.
432
14.80
A B C D E F G
29 ANOVA
31 Sample 211.78 2 105.89 21.04 0.0000 3.22
32 Columns 0.59 1 0.59 0.12 0.7348 4.07
33 Interaction 0.13 2 0.0640 0.0127 0.9874 3.22
34 Within 211.42 42 5.03
35
36 Total 423.91 47
The p-values for interaction, methods, and skills are .9874, .7348, 0, and. The only source of
variation is skill level.
14.81a The factors are mental outlook (2 levels) and physical condition (3 levels).
A B C D E F G
29 ANOVA
31 Sample 2118.4 2 1059.22 24.48 0.0000 3.17
32 Columns 166.7 1 166.67 3.85 0.0548 4.02
33 Interaction 20.0 2 10.02 0.23 0.7941 3.17
34 Within 2336.2 54 43.26
35
36 Total 4641.3 59
Test for interaction: F = .23, p-value = .7941. There is not enough evidence to infer interaction.
b F = 3.85, p-value = .0548. There is not enough evidence to conclude that differences exist
between optimists and pessimists.
c F = 24.48, p-value = 0. There is sufficient evidence to conclude that differences exist between
the three levels of physical condition.
14.82a H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
11 ANOVA
13 Between Groups 9.90 3 3.30 7.67 0.0001 2.70
14 Within Groups 41.33 96 0.43
15
16 Total 51.23 99
F = 7.67, p-value = .0001. There is sufficient evidence to infer that differences in productivity
exist between the four groups of companies.
433
b
A B C D E
1 Multiple Comparisons
2
3 LSD Omega
4 Treatment Treatment Difference Alpha = 0.0083 Alpha = 0.05
5 Extensive Some 0.534 0.500 0.483
6 Little 0.722 0.500 0.483
7 No 0.811 0.500 0.483
8 Some Little 0.188 0.500 0.483
9 No 0.277 0.500 0.483
10 Little No 0.089 0.500 0.483
Using either the Bonferroni adjustment or Tukey’s method we conclude that µ1 differs from µ 2 ,
µ 3 and µ 4 . Companies that offered extensive training have productivity levels different from the
other companies.
14.83a H 0 : µ1 = µ 2 = µ 3 = µ 4 = µ 5
A B C D E F G
12 ANOVA
14 Between Groups 1747.4 4 436.86 4.46 0.0017 2.41
15 Within Groups 23983.7 245 97.89
16
17 Total 25731.1 249
F = 4.46, p-value = .0017. There is enough evidence to infer that there are differences in the effect
of the new assessment system between the five boroughs.
434
b
Multiple Comparisons
LSD Omega
Treatment Treatment Difference Alpha = 0.05 Alpha = 0.05
Borough A Borough B -7.42 3.90 5.40
Borough C -6.18 3.90 5.40
Borough D -2.42 3.90 5.40
Borough E -3.94 3.90 5.40
Borough B Borough C 1.24 3.90 5.40
Borough D 5.00 3.90 5.40
Borough E 3.48 3.90 5.40
Borough C Borough D 3.76 3.90 5.40
Borough E 2.24 3.90 5.40
Borough D Borough E -1.52 3.90 5.40
The mean assessments in borough A differs from the means in boroughs B and C.
c The assessments for each borough are required to be normally distributed with equal variances.
d The histograms are approximately bell-shaped and the sample variances are similar.
14.84 H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
31 ANOVA
33 Rows 43980 19 2314.72 21.58 0.0000 1.77
34 Columns 4438 3 1479.21 13.79 0.0000 2.77
35 Error 6113 57 107.25
36
37 Total 54530 79
F = 13.79, p-value = 0. There is sufficient evidence to conclude that the reading speeds differ
between the four typefaces. The typeface that was read the fastest should be used.
14.85 H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 406.5 2 203.25 1.82 0.1662 3.06
13 Within Groups 16445.8 147 111.88
14
15 Total 16852.3 149
F = 1.82, p-value = .1662. There is not enough evidence to infer that differences in attention span
exist between the three products.
435
14.86 H 0 : µ1 = µ 2 = µ 3
H1 : At least two means differ
A B C D E F G
17 ANOVA
19 Rows 195.33 6 32.56 11.55 0.0002 3.00
20 Columns 43.52 2 21.76 7.72 0.0070 3.89
21 Error 33.81 12 2.82
22
23 Total 272.67 20
F = 7.72, p-value = .0070. There is enough evidence to infer that differences in attention span exist
between the three products.
14.87
A B C D E F G
23 ANOVA
25 Sample 123553 1 123553 3.66 0.0576 3.91
26 Columns 3965110 2 1982555 58.78 0.0000 3.06
27 Interaction 30006 2 15003 0.44 0.6418 3.06
28 Within 4856578 144 33726
29
30 Total 8975248 149
Interaction: F = .44, p-value = .6418. There is not enough evidence to conclude that age and
gender interact.
Age: F = 58.78, p-value = 0. There is sufficient evidence to conclude that differences in offers
exist between the three age groups.
Gender: F = 3.66, p-value = .0576. There is not enough evidence to conclude that differences in
offers exist between males and females
14.88a H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 1769.5 2 884.74 136.58 0.0000 3.02
13 Within Groups 2409.8 372 6.48
14
15 Total 4179.3 374
F = 136.58, p-value = 0. There is sufficient evidence to infer that differences exist between the
effects of the three teaching approaches.
436
b
A B C D E
2
3 LSD Omega
5
Whole Language
Embedded -0.856 0.774 0.754
6 Pure -4.976 0.774 0.754
7 Embedded Pure -4.120 0.774 0.754
All three means differ from one another. From the sample means we may infer that the pure
method is best, followed by embedded, and by whole-language.
14.89a H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 1913 2 956.70 9.54 0.0002 3.10
13 Within Groups 8727 87 100.31
14
15 Total 10640 89
F = 9.54, p-value = .0002. There is sufficient evidence to infer that there are differences between
the three groups.
b
A B C D E
2
3 LSD Omega
5 Mozart White noise -9.30 6.31 6.14
6 Glass -10.20 6.31 6.14
7 White noise Glass -0.90 6.31 6.14
The mean time of the Mozart group differs from the mean times of white noise and the Glass
groups.
14.90 H 0 : µ1 = µ 2 = µ 3 = µ 4
437
A B C D E F G
11 ANOVA
13 Between Groups 5990284 3 1996761 14.47 0.0000 2.64
14 Within Groups 40024172 290 138014
15
16 Total 46014456 293
F = 14.47, p-value = 0. There is enough evidence to infer differences in debt levels between the
four types of degrees.
14.91 H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
11 ANOVA
13 Between Groups 3263 3 1087.8 10.26 0.0000 2.64
14 Within Groups 29685 280 106.0
15
16 Total 32948 283
F = 10.26, p-value = 0. There is enough evidence of differences between the four groups of
investors.
14.92 H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
11 ANOVA
13 Between Groups 3007 3 1002.3 13.84 0.0000 2.67
14 Within Groups 10576 146 72.4
15
16 Total 13583 149
F = 13.84, p-value = 0. There is enough evidence to infer that the length of time depends on the
size of the party
14.93 H 0 : µ1 = µ 2 = µ 3 = µ 4
438
A B C D E F G
11 ANOVA
13 Between Groups 2.12 3 0.705 9.17 0.0000 2.69
14 Within Groups 7.99 104 0.0769
15
16 Total 10.11 107
F = 9.17, p-value = 0. There is sufficient evidence to infer that there are differences in changes to
the TSE depending on the loss the previous day.
14.94 H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 1.57 2 0.787 1.62 0.202233 3.09
13 Within Groups 46.98 97 0.484
14
15 Total 48.55 99
F = 1.62, p-value = .2022. There is no evidence to infer that at least one buy indicator is useful.
14.95 H 0 : µ1 = µ 2 = µ 3
F = 25.98, p-value ≈ 0. There is enough evidence to infer that the amount of sleep differs between
commuting categories.
439
14.96 H 0 : µ1 = µ 2 = µ 3
A B C D E F G
1 ANOVA
3 Between Groups 11374 2 5687 45.49 0.0000 3.00
4 Within Groups 229170 1833 125.0
5
6 Total 240544 1835
F = 45.49, p-value = 0. There is enough evidence to infer that family incomes differ between the
three market segments.
14.97 H 0 : µ1 = µ 2 = µ 3
A B C D E F G
1 ANOVA
3 Between Groups 52.82 2 26.41 3.24 0.0414 3.04
4 Within Groups 1607.82 197 8.16
5
6 Total 1660.6 199
F = 3.24, p-value = .0414. There is enough evidence to infer that the distances driven differ
between drivers who have had 0, 1, or 2 accidents.
14.98 H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
1 ANOVA
3 Between Groups 6636.1 3 2212.03 211.61 0.0000 2.63
4 Within Groups 3595.9 344 10.45
5
6 Total 10232.0 347
F = 211.61, p-value = 0. There is enough evidence to conclude that there are differences in the age
of the car between the four market segments.
440
Case 14.1
Episodes
A B C D E F G
1 Anova: Single Factor
2
3 SUMMARY
4 Groups Count Sum Average Variance
5 Surgery 60 198 3.30 1.74
6 Drug 60 178 2.97 1.19
7 Placebo 60 207 3.45 1.68
8
9
10 ANOVA
12 Between Groups 7.34 2 3.67 2.40 0.0941 3.0470
13 Within Groups 271.4 177 1.53
14
15 Total 278.7 179
F = 2.40, p-value = .0941. There is not enough evidence to conclude that there are differences in
the number of episodes between the three types of treatments.
Visits
A B C D E F G
2
3 SUMMARY
5 Surgery 60 130 2.17 2.51
6 Drug 60 114 1.90 1.38
7 Placebo 60 147 2.45 2.83
8
9
10 ANOVA
12 Between Groups 9.08 2 4.54 2.03 0.1349 3.0470
13 Within Groups 396.6 177 2.24
14
15 Total 405.7 179
the number of physician visits between the three types of treatments.
441
Prescriptions
A B C D E F G
2
3 SUMMARY
5 Surgery 60 201 3.35 2.20
6 Drug 60 178 2.97 3.05
7 Placebo 60 205 3.42 3.43
8
9
10 ANOVA
12 Between Groups 7.08 2 3.54 1.22 0.2968 3.0470
13 Within Groups 512.2 177 2.89
14
15 Total 519.2 179
the number of prescriptions between the three types of treatments.
Days
A B C D E F G
2
3 SUMMARY
5 Surgery 60 689 11.48 22.76
6 Drug 60 663 11.05 19.71
7 Placebo 60 779 12.98 28.22
8
9
10 ANOVA
12 Between Groups 123.5 2 61.76 2.62 0.0756 3.0470
13 Within Groups 4170.8 177 23.56
14
15 Total 4294.3 179
the number of days with respiratory infections between the three types of treatments.
442
Appendix 14
A14.1 One-way Analysis of variance

H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 91.43 2 45.72 1.09 0.3441 3.16
13 Within Groups 2397.5 57 42.06
14
15 Total 2488.9 59
F = 1.09, p-value = .3441. There is no evidence to infer that sales of candy differ according to
placement.
H0 : µD = 0
H1 : µ D < 0
A B C
2
3 Price shown Price not shown
4 Mean 56.15 60.31
5 Variance 243.68 467.71
9 df 99
10 t Stat -3.12
11 P(T<=t) one-tail 0.0012
13 P(T<=t) two-tail 0.0024
t = –3.12, p-value = .0012. There is overwhelming evidence to conclude that ads with no price
shown are more effective in generating interest than ads that show the price.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
442
A B C
2
3 Leftover Returned
4 Mean 61.71 70.57
5 Variance 48.99 203.98
8 df 44
9 t Stat -3.27
10 P(T<=t) one-tail 0.0011
12 P(T<=t) two-tail 0.0021
t = –3.27, p-value = .0011. There is enough evidence to support the professor's theory.
A14.4 a z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
A B C D
2
3 Topiramate Placebo
7 z Stat 1.76
10 P(Z<=z) two-tail 0.0780
z = 1.76, p-value = .0390. There is enough evidence to conclude that topiramate is effective in
causing abstinence for the first month.
b z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
443
A B C D
2
3 Topiramate Placebo
7 z Stat 3.64
10 P(Z<=z) two-tail 0.0002
z = 3.64, p-value = .0001. There is enough evidence to conclude that topiramate is effective in
causing alcoholics to refrain from binge drinking in the final month.
A14.5a Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Expenses MSA Expenses Regular
4 Mean 347.24 479.25
5 Variance 21043 21128
9 df 202
10 t Stat -6.00
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = –6.00, p-value = 0. There is enough evidence to infer that medical expenses for those under the
MSA plan are lower than those who are not.
b z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) < 0
444
A B C D
2
3 Health MSA Health Regular
7 z Stat -0.29
10 P(Z<=z) two-tail 0.7734
z = –.29, p-value = .3867. There is not enough evidence to support the critics of MSA.
A14.6 a One-way analysis of variance

A B C D E F G
11 ANOVA
13 Between Groups 25113 3 8371.2 85.98 0.0000 2.63
14 Within Groups 30766 316 97.36
15
16 Total 55880 319
F = 85.98, p-value = 0.
Two-factor analysis of variance

A B C D E F G
23 ANOVA
25 Sample 17024 1 17024 174.85 0.0000 3.87
26 Columns 7411 1 7411 76.12 0.0000 3.87
27 Interaction 679 1 678.6 6.97 0.0087 3.87
28 Within 30766 316 97.36
29
30 Total 55880 319
Interaction: F = 6.97; p-value = .0087. There is enough evidence to infer that differences are
caused by interaction. There is no need to conduct the other two tests.
A14.7 t-tests of µ
45 minutes: H 0 : µ = 45
H1 : µ < 45
445
A B C D
1 t-Test: Mean
2
3 45 minutes
4 Mean 41.75
7 df 19
8 t Stat -4.01
11 P(T<=t) two-tail 0.0008
60 minutes: H 0 : µ = 60
H1 : µ < 60
A B C D
1 t-Test: Mean
2
3 60 minutes
4 Mean 58.75
7 df 19
8 t Stat -1.11
11 P(T<=t) two-tail 0.2798
80 minutes: H 0 : µ = 80
H1 : µ < 80
A B C D
1 t-Test: Mean
2
3 80 minutes
4 Mean 69.05
7 df 19
8 t Stat -7.76
11 P(T<=t) two-tail 0.0000
446
100 minutes: H 0 : µ = 100
H1 : µ < 100
A B C D
1 t-Test: Mean
2
3 100 minutes
4 Mean 90.40
7 df 19
8 t Stat -3.48
11 P(T<=t) two-tail 0.0026
125 minutes: H 0 : µ = 125
H1 : µ < 125
A B C D
1 t-Test: Mean
2
3 125 minutes
4 Mean 110.05
7 df 19
8 t Stat -3.91
11 P(T<=t) two-tail 0.0010
Overall Conclusion: p-values are .0004, .1399, 0, .0013, and .0005, respectively. In four of the
jobs there is overwhelming evidence to conclude that the times specified by the schedule are
greater than the actual times.
A14.8 Two-way analysis of variance

H 0 : µ1 = µ 2 = µ 3
447
A B C D E F G
35 ANOVA
37 Rows 31,154,590 24 1,298,108 15.05 5.20E-15 1.75
38 Columns 913,217 2 456,608 5.29 0.0084 3.19
39 Error 4,141,276 48 86,277
40
41 Total 36,209,083 74
F = 5.29; p-value = .0084. There is sufficient evidence to infer that differences exist between the
estimated repair costs from different appraisers.
A14.9 One-way analysis of variance

H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 126.0 2 63.02 3.30 0.0439 3.16
13 Within Groups 1087.3 57 19.08
14
15 Total 1213.3 59
F = 3.30, p-value = .0439. There is evidence to infer that at least one rust-proofing method is
different from the others.
A14.10 z-test of p1 − p 2 (case 2)
H0: (p1 – p2) = -.15

H1: (p1 – p2) < - .15
A B C
2
3 Comm 1 Comm 2
6 Hypothesized Difference -0.15
7 z Stat -2.28
10 P(Z<=z) two-tail 0.0228
z = –2.28, p-value = .0114. There is evidence to indicate that the second commercial is viable.
448
H0 : µD = 0
H1 : µ D < 0
A B C
2
3 Prior After
4 Mean 24.91 26.24
5 Variance 48.65 87.88
9 df 99
10 t Stat -2.29
11 P(T<=t) one-tail 0.0121
13 P(T<=t) two-tail 0.0242
t = −2.29, p-value = .0121. There is enough evidence to conclude that company should proceed to
stage 2.
A14.12 Two-factor analysis of variance

A B C D E F G
29 ANOVA
31 Sample 427.61 2 213.81 39.97 0.0000 3.06
32 Columns 20.17 1 20.17 3.77 0.0541 3.91
33 Interaction 17.77 2 8.89 1.66 0.1935 3.06
34 Within 770.32 144 5.35
35
36 Total 1235.87 149
Interaction: F = 1.66, p-value = .1935. There is no evidence of interaction.
Gender (Columns) : F = 3.77, p-value = .0541. There is not enough evidence of a difference
between men and women.
Fitness (Sample): F = 39.97, p-value = 0. There is overwhelming evidence of differences among
the three levels of fitness.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
449
A B C
2
3 ABS speed No ABS speed
4 Mean 34.72 33.94
5 Variance 25.27 25.63
9 df 198
10 t Stat 1.09
11 P(T<=t) one-tail 0.1394
13 P(T<=t) two-tail 0.2788
t = 1.09, p-value = .2788. There is not enough evidence that operating an ABS-equipped car
changes a driver's behavior.

A B C D E F G
10 ANOVA
12 Between Groups 1813.7 2 906.87 6.46 0.0030 3.16
13 Within Groups 7998.0 57 140.32
14
15 Total 9811.7 59
Multiple comparisons
A B C D E
2
3 LSD Omega
5 Price: $34 Price: $39 -12.9 9.24 9.01
6 Price: $44 -3.1 9.24 9.01
7 Price: $39 Price: $44 9.8 9.24 9.01
Sales with $34 and $44 dollar prices do not differ. Sales with $39 differ from sales with $34 and
$44 prices.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
450
A B C
2
3 Echinacea Placebo
4 Mean 7.02 7.06
9 df 522
10 t Stat -0.31
11 P(T<=t) one-tail 0.3799
13 P(T<=t) two-tail 0.7598
t = –.31, p-value = .3799. There is not enough evidence to conclude that Echinacea is effective.

A B C D E F G
24 ANOVA
26 Rows 335.17 13 25.78 16.86 0.0000 2.12
27 Columns 10.90 2 5.45 3.57 0.0428 3.37
28 Error 39.76 26 1.53
29
30 Total 385.83 41
a H 0 : µ1 = µ 2 = µ 3
F = 3.57, p-value = .0428. There is enough evidence to conclude that there are differences in
waiting times between the three resorts.
b The waiting times are required to be normally distributed with the same variance at all three
resorts.
c Histograms are used to check the normality requirement.
A14.17a There are 4 levels of ranks and 4 levels of faculties for a total of 16 treatments.
b
A B C D E F G
23 ANOVA
25 Between Groups 1091.15 15 72.74 2.84 0.0019 1.83
26 Within Groups 1638.40 64 25.60
27
28 Total 2729.55 79
451
F = 2.84, p-value = .0019. There is enough evidence to infer that at least two treatment means
differ.
c Factor A (columns) is the faculty. The levels are business, engineering, arts, and science. Factor
B (samples) is the rank. The levels are professor, associate professor, assistant professor, and
lecturer.
A B C D E F G
35 ANOVA
37 Sample 46.85 3 15.62 0.61 0.6109 2.75
38 Columns 344.65 3 114.88 4.49 0.0064 2.75
39 Interaction 699.65 9 77.74 3.04 0.0044 2.03
40 Within 1638.40 64 25.60
41
42 Total 2729.55 79
d F = 3.04, p-value = .0044. There is evidence to conclude that ranks and faculties interact.
e F =.61, p-value = .6109. There is not enough evidence to conclude that differences exist between
the ranks. The answer to Part d indicates that this test is irrelevant.
f F = 4.49, p-value = .0064. There is enough evidence to conclude that differences exist between
the faculties. The answer to Part d indicates that this test is irrelevant.
A14.18 a z-estimate of p
A B C D E
2
4 Sample size 1328 0.357 ± 0.026
Estimate of the number of households with at least one dog
LCL = 112 million × .331 = 37.072 million
UCL = 112 million × .383 = 42.896 million
b
A B C D E
2
4 Sample size 1328 0.316 ± 0.025
Number of households with at least one cat
LCL = 112 million × .291 = 32.592 million
UCL = 112 million × .341 = 38.192million
452
c t-estimate of µ
A B C D
1 t-Estimate: Mean
2
3 Dogs
4 Mean 247.19
6 LCL 235.17
7 UCL 259.20
Estimate of the total amount spent on dogs
LCL = 40 million ×235.17 = $9.407 billion
UCL = 40 million ×259.29 = $10.368 billion,
d t-estimate of µ
A B C D
1 t-Estimate: Mean
2
3 Cats
4 Mean 158.07
6 LCL 149.53
7 UCL 166.61
Estimate of the total amount spent on cats
LCL = 35 million ×149.53 = $5.234 billion
UCL = 35 million ×166.61 = $5.831 billion
A14.19 H0: p = .5
H1: p > .5
Success = 1
z = 14.31, p-value = 0. There is enough evidence to conclude that the majority of Americans
support capital punishment.
453
A14.20 H0: (p1 – p2) = 0
H1: (p1 – p2) ≠ 0
z = .918, p-value = .3588. There is not enough evidence to conclude that Democrats and
Republicans differ in their responses to the question about the use of any drugs.

H 0 : µ1 = µ 2 = µ 3 = µ 4
F = 12.83, p-value = 2.89E-08 ≈ 0. There is enough evidence to infer that differences in the
amount of television watched exist between the four classes.
454
H 0 : µ1 = µ 2 = µ 3
F = .88, p-value = .4149. There is not enough evidence to conclude that the number of hours
worked differ between the races.
A14.23 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
Two-tail F test: F = 1.05, p-value = .1570, Use equal-variances t- test
t = 1.24, p-value = .1074. There is not enough evidence to infer that the average age of Americans
in 2008 is larger than the average in 2006.
455
A14.24 H0: (p1 – p2) = 0
H1: (p1 – p2) ≠ 0
z = -5.55, p-value = 0. There is sufficient evidence to infer that Democrats and Republicans differ
in their access to the Internet.
A14.25 H 0 : µ1 = µ 2 = µ 3
F = 16.98, p-value = 5.26E-08 ≈ 0. There is enough evidence to conclude that liberals, moderates,
and conservatives differ in age.
456
Chapter 15
15.1 H 0 : p1 = .1, p 2 = .2, p 3 = .3, p 4 = .2, p 5 = .2

H1 : At least one p i is not equal to its specified value.
Cell i fi ei (f i − e i ) (f i − e i ) 2 / e i
1 24 300(.1) = 30 -6 1.20
2 64 300(.2) = 60 4 .27
3 84 300(.3) = 90 -6 .40
4 72 300(.2) = 60 12 2.40
5 56 300(.2) = 60 -4 .27
Total 300 300 χ 2 = 4.54
Rejection region: χ 2 > χ α2 ,k −1 = χ .201,4 = 13.3
χ 2 = 4.54, p-value = .3386. There is not enough evidence to infer that at least one p i is not equal
to its specified value.
15.2 H 0 : p1 = .1, p 2 = .2, p 3 = .3, p 4 = .2, p 5 = .2

1 12 150(.1) = 15 -3 .60
2 32 150(.2) = 30 2 .13
3 42 150(.3) = 45 -3 .20
4 36 150(.2) = 30 6 1.20
5 28 150(.2) = 30 -2 .13
Total 150 150 χ 2 = 2.26
Rejection region: χ 2 > χα2 ,k −1 = χ.201, 4 = 13.3
15.3 H 0 : p1 = .1, p 2 = .2, p 3 = .3, p 4 = .2, p 5 = .2

457
1 6 75(.1) = 7.5 -1.5 .30
2 16 75(.2) = 15 1 .07
3 21 75(.3) = 22.5 -1.5 .10
4 18 75(.2) = 15 3 .60
5 14 70(.2) = 15 -1 .07
Total 75 75 χ 2 = 1.14
Rejection region: χ 2 > χα2 ,k −1 = χ.201, 4 = 13.3
15.4 The χ 2 statistic decreases.
15.5 H 0 : p1 = .3, p 2 = .3, p 3 = .2, p 4 = .2

1 38 150(.3) = 45 -7 1.09
2 50 150(.3) = 45 5 0.56
3 38 150(.2) = 30 8 2.13
4 24 150(.2) = 30 -6 1.20
Total 150 150 χ 2 = 4.98
Rejection region: χ 2 > χα2 ,k −1 = χ.205,3 = 7.81
15.6 H 0 : p1 = .3, p 2 = .3, p 3 = .2, p 4 = .2

1 76 300(.3) = 90 -14 2.18
2 100 300(.3) = 90 10 1.11
3 76 300(.2) = 60 16 4.27
4 48 300(.2) = 60 -12 2.40
Total 300 300 χ 2 = 9.96
458
χ 2 = 9.96, p-value = .0189. There is enough evidence to infer that at least one p i is not equal to its
specified value.
15.7 H 0 : p1 = .2, p 2 = .2, p 3 = .2, p 4 = .2, p 5 = .2

1 28 100(.2) = 20 8 3.20
2 17 100(.2) = 20 -3 0.45
3 19 100(.2) = 20 -1 0.05
4 17 100(.2) = 20 -3 0.45
5 19 100(.2) = 20 -1 0.05
Total 100 100 χ 2 = 4.20
15.8 H 0 : p1 = .15, p 2 = .40, p 3 = .35, p 4 = .10

1 41 233(.15) = 34.95 6.05 1.05
2 107 233(.40) = 93.20 13.80 2.04
3 66 233(.35) = 81.55 -15.55 2.97
4 19 233(.10) = 23.30 -4.30 0.79
2
Total 233 233 χ = 6.85
15.9 H 0 : p1 = 1/6, p 2 = 1/6, p 3 = 1/6, p 4 = 1/6, p 5 = 1/6, p 6 = 1/6

459
1 114 600(1/6) = 100 14 1.96
2 92 600(1/6) = 100 -8 0.64
3 84 600(1/6) = 100 -16 2.56
4 101 600(1/6) = 100 1 0.01
5 107 600(1/6) = 100 7 0.49
6 102 600(1/6) = 100 2 0.04
Total 600 600 χ 2 = 5.70
χ 2 = 5.70, p-value = .3365. There is not enough evidence to infer that the die is not fair.
15.10 H 0 : p1 = .05, p 2 = .25 p 3 = .40, p 4 = .25 p 5 = .05

1 11 150(.05) = 7.5 3.5 1.63
2 32 150(.25) = 37.5 -5.5 0.81
3 62 150(.40) = 60.0 2.0 0.07
4 29 150(.25) = 37.5 -8.5 1.93
5 16 150(.05) = 7.5 8.5 9.63
Total 150 150 χ 2 = 14.07
χ 2 = 14.07, p-value = .0071. There is enough evidence to infer that grades are distributed
differently from grades in the past.
15.11 H 0 : p1 = .2, p 2 = .2 p 3 = .2, p 4 = .2 p 5 = .2

1 8 25(.2) = 5.0 3.0 1.80
2 4 25(.2) = 5.0 -1.0 0.20
3 3 25(.2) = 5.0 -2.0 0.80
4 8 25(.2) = 5.0 3.0 1.80
5 2 25(.2) = 5.0 -3.0 1.80
Total 25 25 χ 2 = 6.40
460
χ 2 = 6.40, p-value = .1712. There is not enough evidence to infer that the professor does not
randomly distribute the correct answer over the five choices.
15.12 H 0 : p1 = .72, p 2 = .15, p 3 = .10, p 4 = .03

1 159 250(.72) = 180.0 -21.0 2.45
2 28 250(.15) = 37.5 -9.5 2.41
3 47 250(.10) = 25.0 22.0 19.36
4 16 250(.03) = 7.5 8.5 9.63
Total 250 250 χ 2 = 33.85
χ 2 =33.85, p-value = 0. There is enough evidence to infer that the aging schedule has changed.
15.13 H 0 : p1 = .15, p 2 = .25, p 3 = .40, p 4 = .20

1 36 250(.15) = 29.55 6.45 1.41
2 58 250(.25) = 49.25 8.75 1.55
3 74 250(.40) = 78.80 -4.80 0.29
4 29 250(.20) = 39.40 -10.40 2.75
Total 197 197 χ 2 = 6.00
χ 2 = 6.00, p-value = .1116. There is not enough evidence to infer that certain sizes of cars are
involved in a higher than expected percentage of accidents.
15.14 H 0 : p1 = .31, p 2 = .51, p 3 = .18

461
1 408 1200(.31) = 372 36 3.48
2 571 1200(.51) = 612 -41 2.75
3 221 1200(.18) = 216 5 0.12
Total 1200 1200 χ 2 = 6.35
Rejection region: χ 2 > χ α2 ,k −1 = χ.210,2 = 4.61
χ 2 = 6.35, p-value = .0419. There is enough evidence to infer that voter support has changed since
the election.
15.15 H 0 : p1 = .05, p 2 = .07, p 3 = .04, p 4 = .84

1 19 250(.05) = 12.5 6.5 3.38
2 23 250(.07) = 17.5 5.5 1.73
3 14 250(.04) = 10.0 4.0 1.60
4 194 250(.84) = 210.0 -16.0 1.22
Total 250 250 χ 2 = 7.93
χ 2 = 7.93, p-value = .0475. There is enough evidence to infer that the reported side effects of the
placebo differ from that of the cold remedy.
15.16 H 0 : p1 = .23, p 2 = .40, p 3 = .15, p 4 = .22

1 63 320(.23) = 73.6 -10.6 1.53
2 125 320(.40) = 128.0 -3.0 0.07
3 45 320(.15) = 48.0 -3.0 0.19
4 87 320(.22) = 70.4 16.6 3.91
2
Total 320 320 χ = 5.70
χ 2 = 5.70, p-value = .1272. There is not enough evidence to infer that there has been a change in
proportions.
462
15.17 H0: p1 = .58, p2 = .06, p3 = .11, p4 = .25
H1: At least one pi is not equal to its specified value.
Marital status Observed Expected
Married & separated 794 + 179 = 973 .58(1782) = 1033.6
Widowed 281 .06(1782) = 106.9
Divorced 66 .11(1782) = 196.0
Never married & partnered 432 + 30 = 462 .25(1782) = 445.5
Total 1782
χ 2 = 373.9, p-value = 9.80E-81 ≈ 0. There is sufficient evidence to conclude that the ANES in
2008 over represented at least one category of marital status.
15.18 H0: p1 = .79, p2 = .13, p3 = .08

Race Observed Expected
White 1559 .79(2023) = 1598.2
Black 281 .13(2023) = 263.0
Other 183 .08(2023) = 161.8
Total 2023
2
χ = 4.97, p-value = .0833. There is not enough evidence to infer that the GSS in 2008
overrepresented at least one race.
15.19 H0: p1 = .79, p2 = .13, p3 = .08

Race Observed Expected
White 3284 .79(4510) = 3562.9
Black 634 .13(4510) = 586.3
Other 592 .08(4510) = 360.8
Total 4510
χ 2 =173.9 p-value =1.76E-38 ≈ 0.. There is enough evidence to infer that the GSS in 2006
overrepresented at least one race.
15.20 H0: p1 = .58, p2 = .06, p3 = .11, p4 = .25

463
Married & separated 972 + 70 = 1042 .58(2018) = 1170.4
Widowed 164 .06(2018) = 121.1
Divorced 281 .11(2018) = 222.0
Never married 531 .25(2018) = 504.5
Total 2018
χ 2 = 46.36, p-value = 4.76E-10 ≈ 0. There is sufficient evidence to conclude that the GSS in 2008
over represented at least one category of marital status.
15.21 H0: p1 = .58, p2 = .06, p3 = .11, p4 = .25

Married & separated 2170+156 = 2326 .58(4504) = 2612.3
Widowed 366 .06(4504) = 270.2
Divorced 732 .11(4504) = 495.4
Never married 1080 .25(4504) = 1126
Total 4504
χ 2 = 180.2, p-value = 7.90E-39 ≈ 0. There is sufficient evidence to conclude that the GSS in 2006
over represented at least one category of marital status.
15.22 H 0 : The two variables are independent

H1 : The two variables are dependent
1 28 96(84)/188 = 42.89 -14.89 5.17
2 68 96(104)/188 = 53.11 14.89 4.17
3 56 92(84)/188 = 41.11 14.89 5.40
4 36 92(104)/188 = 50.89 -14.89 4.36
Total 188 188 χ 2 = 19.10
Rejection region: χ 2 > χ α2 ,( r −1)( c −1 = χ .205,1 = 3.84
χ 2 = 19.10, p-value = 0. There is enough evidence to infer that the two variables are dependent.

464
1 14 48(42)/188 = 21.45 -7.45 2.59
2 34 48(52)/188 = 26.55 7.45 2.09
3 28 46(42)/188 = 20.55 7.45 2.70
4 18 46(52)/188 = 25.45 -7.45 2.18
Total 94 94 χ 2 = 9.56
χ 2 = 9.56, p-value = .0020. There is enough evidence to infer that the two classifications L and M
are dependent.

1 7 24(21)/188 = 10.72 -3.72 1.29
2 17 24(26)/188 = 13.28 3.72 1.04
3 14 23(21)/188 = 10.28 3.72 1.35
4 9 23(26)/188 = 12.72 -3.72 1.09
Total 47 47 χ 2 = 4.77
χ 2 = 4.77, p-value = .0289. There is enough evidence to infer that the two classifications L and M
are dependent.
15.25 The χ 2 statistic decreases.

1 40 120(70)/250 = 33.60 6.40 1.22
2 32 120(80)/250 = 38.40 - 6.40 1.07
3 48 120(100)/250 = 48.00 0 0.00
4 30 130(70)/250 = 36.40 -6.40 1.13
5 48 130(80)/250 = 41.60 6.40 0.99
6 52 130(100)/250 = 52.00 0 0.00
465
Total 250 250 χ 2 = 4.41
Rejection region: χ 2 > χ α2 ,( r −1)( c −1 = χ .210, 2 = 4.61
χ 2 = 4.41, p-value = .1110. There is not enough evidence to infer that the two classifications R
and C are dependent.
15.27 H 0 : The two variables (responses and employee group) are independent
1 67 110(130)/200 = 71.50 -4.50 0.28
2 32 110(50)/200 = 27.50 4.50 0.74
3 11 110(20)/200 = 11.00 0 0.00
4 63 90(130)/200 = 58.50 4.50 0.35
5 18 90(50)/200 = 22.50 -4.50 0.90
6 9 90(20)/200 = 9.00 0 0.00
2
Total 200 200 χ = 2.27
χ 2 = 2.27, p-value = .3221. There is not enough evidence to infer that responses differ among the
three groups of employees.
15.28 H 0 : The two variables (shirt condition and shift) are independent
1 240 570(250)/600 = 237.5 2.5 .03
2 191 570(200)/600 = 190.0 1.0 .01
3 139 570(150)/600 = 142.5 -3.5 .09
4 10 30(250)/600 = 12.5 -2.5 .50
5 9 30(200)/600 = 10.0 -1.0 10
6 11 30(150)/600 = 7.5 3.5 1.63
Total 600 600 χ 2 = 2.36
χ 2 = 2.36, p-value = .3087. There is not enough evidence to infer that there are differences in
quality among the three shifts.
466
15.29 H 0 : The two variables economic option and political affiliation) are independent
1 101 444(331)/1000 = 146.96 -45.96 14.376
2 282 444(557)/1000 = 233.99 48.01 9.852
3 61 444(142)/1000 = 63.05 -2.05 0.067
4 38 130(331)/1000 = 43.03 -5.03 0.588
5 67 130(557)/1000 = 68.51 -1.51 0.033
6 25 130(142)/1000 = 18.46 6.54 2.317
7 131 250(331)/1000 = 82.75 48.25 28.134
8 88 250(557)/1000 = 131.75 -43.75 14.528
9 31 250(142)/1000 = 35.50 -4.50 0.570
10 61 176(331)/1000 = 58.26 2.74 0.129
11 90 176(557)/1000 = 92.75 -2.75 0.082
12 25 176(142)/1000 = 24.99 0.01 0.000
Total 1000 1000 χ 2 = 70.675
χ 2 = 70.675, p-value = 0. There is sufficient evidence to infer that political affiliation affects
support for economic options.
15.30 H 0 : The two variables (inducement and return) are independent

1 80 300(200)/1000 = 60 20 6.67
2 100 300(300)/1000 = 90 10 1.11
3 120 300(500)/1000 = 150 -30 6.00
4 120 700(200)/1000 = 140 -20 2.86
5 200 700(300)/1000 = 210 -10 0.50
6 380 700(500)/1000 = 350 30 2.57
Total 1000 1000 χ 2 = 19.71
Rejection region: χ 2> χ α2 ,( r −1)( c−1 = χ .205, 2 = 5.99
χ 2 = 19.71, p-value = .0001. There is sufficient evidence to infer that the return rates differ among
the different inducements.
467
15.31 H 0 : The two variables (newspaper and occupation) are independent
1 27 120(89)/ 354=30.2 -3,2 .33
2 18 120(112)/354=38.0 -20.0 10.50
3 38 120(81)/354=27.5 10.5 4.05
4 37 120(72)/354=24.4 12.6 6.50
5 29 108(89)/354=27.2 1.8 .13
6 43 108(112)/354=34.2 8.8 2.28
7 21 108(81)/354=24.7 -3.7 .56
8 15 108(72)/354=22.0 -7.0 2.21
9 33 126(89)/354=31.7 1.3 .06
10 51 126(112)/354=39.9 11.1 3.11
11 22 126(81)/354=28.8 -6.8 1.62
12 20 126(72)/354=25.6 -5.6 1.24
Total 354 354 χ 2 = 32.57
Rejection region: χ 2> χ α2 ,( r −1)( c −1 = χ .205,6 = 12.6
χ 2 = 32.57, p-value = 0. There is sufficient evidence to infer that occupation and newspaper are
related.
15.32a H 0 : The two variables (predicted change and actual change) are independent
1 65 129(104)/216 = 62.11 2.89 .13
2 39 87(104)/216 = 41.89 -2.89 .20
3 64 129(112)/216 =66.89 -2.89 .12
4 48 87(112)/216 = 45.11 2.89 .19
Total 216 216 χ 2 = .64
χ 2 = .64, p-value = .4225. There is not enough evidence to infer that the predicted and actual
directions of change are related.

b Ignore what the other investors are doing.
468
15.33 H 0 : The two variables (last purchase and second-last purchase) are independent
1 39 149(153)/559 = 40.78 −1.78 .08
2 36 149(134)/559 = 35.72 .28 0
3 51 149(190)/559 = 50.64 .36 0
4 23 149(82)/559 = 21.86 1.14 .06
5 36 134(153)/559 = 36.68 −.68 .01
6 32 134(134)/559 = 32.12 −.12 0
7 46 134(190)/559 = 45.55 .45 0
8 20 134(82)/559 = 19.66 .34 .01
9 54 194(153)/559 = 53.10 .90 .02
10 46 194(134)/559 = 46.50 −.50 .01
11 65 194(190)/559 = 65.94 −.94 .01
12 29 194(82)/559 = 28.46 .54 .01
13 24 82(153)/559 = 22.44 1.56 .11
14 20 82(134)/559 = 19.66 .34 .01
15 28 82(190)/559 = 27.87 .13 0
16 10 82(82)/558 = 12.03 −2.03 .34
2
Total 559 559 χ = .67
Rejection region: χ 2> χ α2 ,( r −1)( c −1 = χ .205,9 = 16.9
χ 2 = .67, p-value = .9999. There is no evidence of a relationship.
15.34 H 0 : The two variables (education and smoker) are independent

1 34 57(460)/1000 = 26.22 7.78 2.31
2 23 57(540)/1000 = 30.78 -7.78 1.97
3 251 463(460)/1000 = 212.98 38.02 6.79
4 212 463(540)/1000 = 250.02 -38.02 5.78
5 159 407(460)/1000 = 187.22 -28.22 4.25
6 248 407(540)/1000 = 219.78 28.22 3.62
7 16 73(460)/1000 = 33.58 -17.58 9.20
8 57 73(540)/1000 = 39.42 17.58 7.84
Total 1000 1000 χ 2 = 41.77
469
χ 2 = 41.77, p-value = 0. There is sufficient evidence to infer that the amount of education is a
factor in determining whether a smoker will quit.
15.35 H 0 : The two variables (education and smoker) are independent

1 60 121(369)/658 =67.9 -7.9 .91
2 23 121(116)/658=21.3 1.7 .13
3 13 121(65)/658=12.0 1.0 .09
4 25 121(108)/658=19.9 5.1 1.33
5 65 126(369)/658=70.7 -5.7 .45
6 19 126(116)/658=22.2 -3.2 .46
7 14 126(65)/658=12.4 1.6 .19
8 28 126(108)/658=20.7 7.3 2.59
9 73 132(369)/658=74.0 -1.0 .01
10 26 132(116)/658=23.3 2.7 .32
11 9 132(65)/658=13.0 -4.0 1.25
12 24 132(108)/658=21.7 2.3 .25
13 67 95(369)/658=53.3 13.7 3.54
14 11 95(116)/658=16.7 -5.7 1.97
15 10 95(65)/658=9.4 0.6 .04
16 7 95(108)/658=15.6 -8.6 4.74
17 57 96(369)/658=53.8 3.2 .19
18 16 96(116)/658=16.9 -.9 .05
19 9 96(65)/658=9.5 -.5 .02
20 14 96(108)/658=15.8 -1.8 .20
21 47 88(369)/658=49.3 -2.3 .11
22 21 88(116)/658=15.5 5.5 1.94
23 10 88(65)/658=8.7 1.3 .20
24 10 88(108)/658=14.4 -4.4 1.37
2
Total 658 658 χ = 22.36
χ 2 = 22.36, p-value = .0988. There is not enough evidence to infer that there is a relationship
between an adult’s source of news and his or her heartburn condition.
470
15.36 H 0 : The two variables (university and degree) are independent
1 44 100(167)/400=41.75 2.25 .12
2 11 100(64)/ 400=16.00 -5.00 1.56
3 34 100(121)/ 400=30.25 3.75 .46
4 11 100(48)/ 400=12.00 -1.00 .08
5 52 100(167)/ 400=41.75 10.25 2.52
6 14 100(64)/ 400=16.00 -2.00 .25
7 27 100(121)/ 400=30.25 -3.25 .35
8 7 100(48)/ 400=12.00 -5.00 2.08
9 31 100(167)/ 400=41.75 -10.75 2.77
10 27 100(64) 400=16.00 11.00 7.56
11 18 100(121) 400=/30.25 -12.25 4.96
12 24 100(48)/ 400=12.00 12.00 12.00
13 40 100(167)/ 400=41.75 -1.75 .07
14 12 100(64)/ 400=16.00 -4.00 1.00
15 42 100(121)/ 400=30.25 11.75 4.56
16 6 100(49)/ 400=12.00 -6.00 3.00
2
Total 400 400 χ = 43.36
χ 2 = 43.36, p-value = 0. There is enough evidence to infer that undergraduate degree and the
university applied to are related.
15.37 H 0 : The two variables (results and financial ties) are independent
1 29 30(48)/70 = 20.57 8.43 3.45
2 1 30(22)/70 = 9.43 -8.43 7.54
3 10 17(48)/70 = 11.66 -1.66 .24
4 7 17(22)/70 = 5.34 1.66 .52
5 9 23(48)/70 = 15.77 -6.77 2.91
6 14 23(22)/70 =7.23 6.77 6.34
Total 70 70 χ 2 = 21.00
471
χ 2 = 21.00, p-value = 0. There is sufficient evidence to infer that the research findings are related
to whether drug companies fund the research.
15.38 H0: The two variables (degree and approach) are independent
H1: The two variables are dependent
1 51 75(101)/195 = 38.85 12.15 3.80
2 8 75(31)/195 = 11.92 -3.92 1.29
3 5 75(36)/195 = 13.85 -8.85 5.65
4 11 752(27)/195 = 10.38 .62 .04
5 24 58(101)/195 = 30.04 -6.04 1.21
6 14 58(31)/195 = 9.22 4.78 2.48
7 12 58(36)/195 = 10.71 1.29 .16
8 8 58(27)/195 = 8.03 -.03 0
9 26 62(101)/195 =32.11 -6.11 1.16
10 9 62(31)/195 = 9.86 -.86 .07
11 19 62(36)/195 = 11.45 7.55 4.99
12 8 62(27)/195 = 8.58 -.58 .04
Total 195 195 χ 2 = 20.89
Rejection region: χ 2 > χ α2 ,( r −1)( c−1 = χ .205,6 = 12.6
χ 2 = 20.89, p-value = .0019. There is sufficient evidence to infer that there are differences in
teaching approach among the four types of degree. The editor can design books and sales
campaigns based on the distribution of degrees.
15.39 H0: The two variables are independent

472
χ2 = 1.072, p-value = .7838. There is not enough evidence to infer that support for gun laws varied
from year to year.

χ2 = 36.57, p-value = .0003. There is enough evidence to infer that marital status varied from year
to year.

473
χ2 = 184.3, p-value = 0. There is enough evidence to infer that men and women differ in their work
status.

χ2 = 110.35, p-value = 0. There is enough evidence to infer that support for gun laws is related to
political affiliation.

474
χ2 = 31.40, p-value = 0. There is enough evidence to infer that there are differences between the
three party affiliations with respect to knowing where to vote.

χ2 = 5.89, p-value = .0525. There is not enough evidence to infer that there are differences between
the three party affiliations with respect to reading about campaign.

475
three party affiliations with respect to having health insurance.

three party affiliations with respect to access to the Internet.
15.47 H 0 : The data are normally distributed

H1 : The data are not normally distributed
476
Expected Observed
Interval Probability Value e i Value f i f i − ei (f i − ei ) 2 / ei
Z ≤ -1.5 .0668 6.68 10 3.32 1.65
-1.5 < Z ≤ -0.5 .2417 24.17 18 -6.17 1.58
-0.5 < Z ≤ 0.5 .3829 38.29 48 9.71 2.46
0.5 < Z ≤ 1.5 .2417 24.17 16 -8.17 2.76
Z > 1.5 .0668 6.68 8 1.32 0.26
Total 1 100 100 χ 2 = 8.71
χ 2 = 8.71, p-value = .0128. There is enough evidence to infer that the data are not normally
distributed.
15.48 H 0 : The data are normally distributed

H1 : The data are not normally distributed
Expected Observed
Interval Probability Value e i Value f i f i − ei (f i − ei ) 2 / ei
Z ≤ -1 .1587 7.94 6 -1.94 0.47
-1 < Z ≤ 0 .3413 17.07 27 9.93 5.78
0<Z ≤ 1 .3413 17.07 14 -3.07 0.55
Z>1 .1587 7.94 3 -4.94 3.07
Total 1 50 50 χ 2 = 9.87
χ 2 = 9.87, p-value = .0017. There is sufficient evidence to infer that the data are not normally
distributed.
15.49 H 0 : Times are normally distributed

H1 : Times are not normally distributed.
477
A B C D
1 Chi-Squared Test of Normality
2
3 Hours
4 Mean 7.15
5 Standard deviation 1.65
6 Observations 200
7
8 Intervals Probability Expected Observed
9 (z <= -1.5) 0.0668 13.36 11
10 (-1.5 < z <= -0.5) 0.2417 48.35 55
11 (-0.5 < z <= 0.5) 0.3829 76.59 52
12 (0.5 < z <= 1.5) 0.2417 48.35 67
13 (z > 1.5) 0.0668 13.36 15
14
15
17 df 2
18 p-value 0.0002
19 chi-squared Critical 5.9915
χ 2 = 16.62, p-value = .0002. There is sufficient evidence to infer that the amount of time at part-
time jobs is not normally distributed.
15.50 H 0 : Costs are normally distributed

H1 : Costs are not normally distributed
A B C D
2
3 Drug Cost
4 Mean 29.69
6 Observations 900
7
9 (z <= -2) 0.0228 20.48 0
10 (-2 < z <= -1) 0.1359 122.31 9
11 (-1 < z <= 0) 0.3413 307.21 599
12 (0 < z <= 1) 0.3413 307.21 248
13 (1 < z <= 2) 0.1359 122.31 17
14 (z > 2) 0.0228 20.48 27
15
17 df 3
18 p-value 0
χ 2 = 506.76, p-value = 0. There is sufficient evidence to infer that drug costs are not normally
distributed.
478
15.51 Successful firms:
H 0 : Productivity in successful firms is normally distributed
H1 : Productivity in successful firms is not normally distributed
A B C D
2
3 Successful
4 Mean 5.02
6 Observations 200
7
9 (z <= -1.5) 0.0668 13.36 12
10 (-1.5 < z <= -0.5) 0.2417 48.35 52
11 (-0.5 < z <= 0.5) 0.3829 76.59 72
12 (0.5 < z <= 1.5) 0.2417 48.35 55
13 (z > 1.5) 0.0668 13.36 9
14
15
17 df 2
18 p-value 0.2199
χ 2 =3.03, p-value = .2199. There is not enough evidence to infer that productivity in successful
firms is not normally distributed.
Unsuccessful firms:
H 0 : Productivity in unsuccessful firms is normally distributed
H1 : Productivity in unsuccessful firms is not normally distributed
479
A B C D
2
3 Unsuccessful
4 Mean 7.80
6 Observations 200
7
9 (z <= -1.5) 0.0668 13.36 12
10 (-1.5 < z <= -0.5) 0.2417 48.35 47
11 (-0.5 < z <= 0.5) 0.3829 76.59 83
12 (0.5 < z <= 1.5) 0.2417 48.35 44
13 (z > 1.5) 0.0668 13.36 14
14
15
17 df 2
18 p-value 0.567
χ 2 = 1.13, p-value = .5670. There is not enough evidence to infer that productivity in unsuccessful
firms is not normally distributed.
15.52 H 0 : Reaction times are normally distributed

H1 : Reaction times re not normally distributed
Phone
A B C D
2
3 Phone
4 Mean 0.646
6 Observations 125
7
9 (z <= -1.5) 0.0668 8.35 8
10 (-1.5 < z <= -0.5) 0.2417 30.22 32
11 (-0.5 < z <= 0.5) 0.3829 47.87 47
12 (0.5 < z <= 1.5) 0.2417 30.22 29
13 (z > 1.5) 0.0668 8.35 9
14
15
17 df 2
18 p-value 0.8891
480
χ 2 = .2351, p-value = .8891. There is not enough evidence to infer that reaction times of those
using the cell phone are not normally distributed.
Not on phone
A B C D
2
3 Not
4 Mean 0.601
6 Observations 145
7
9 (z <= -1.5) 0.0668 9.69 8
10 (-1.5 < z <= -0.5) 0.2417 35.05 40
11 (-0.5 < z <= 0.5) 0.3829 55.52 55
12 (0.5 < z <= 1.5) 0.2417 35.05 29
13 (z > 1.5) 0.0668 9.69 13
14
15
17 df 2
18 p-value 0.2044
χ 2 = 3.1752, p-value = .2044. There is not enough evidence to infer that reaction times of those
not using the cell phone are not normally distributed.
15.53 H 0 : Matched pairs differences of sales are normally distributed

H1 : Matched pairs differences of sales are not normally distributed
481
A B C D
2
3 Difference
4 Mean 19.75
6 Observations 40
7
9 (z <= -1) 0.1587 6.35 6
10 (-1 < z <= 0) 0.3413 13.65 14
11 (0 < z <= 1) 0.3413 13.65 14
12 (z > 1) 0.1587 6.35 6
13
14
15
17 df 1
18 p-value 0.8140
χ 2 = .055, p-value = .8140. There is not enough evidence to infer that matched pairs difference of
sales is not normally distributed.
15.54 H 0 : p1 = 1/3, p 2 = 1/3, p 3 = 1/3

Cell i fi ei (f i − e i ) (f i − ei ) 2 / ei
1 14 30(1/3) = 10 4 1.60
2 10 30(1/3) = 10 0 0.00
3 6 30(1/3) = 10 -4 1.60
Total 30 30 χ 2 = 3.20
χ 2 = 3.20, p-value = .2019. There is not enough evidence to infer that the game is unfair.
15.55 H 0 : p1 = .2, p 2 = .2, p 3 = .2, p 4 = .2, p 5 = .2

482
1 87 362(.2) = 72.4 14.6 2.94
2 62 362(.2) = 72.4 -10.4 1.49
3 71 362(.2) = 72.4 -1.4 0.03
4 68 362(.2) = 72.4 -4.4 0.27
5 74 362(.2) = 72.4 1.6 0.04
Total 362 362 χ 2 = 4.77
χ 2 = 4.77, p-value = .3119. There is not enough evidence to infer that absenteeism is higher on
some days of the week.
15.56 H 0 : The two variables (shift and day) are independent

1 52 181(87)/362 = 43.50 8.50 1.66
2 28 181(62)/362 = 31.00 -3.00 0.29
3 37 181(71)/362 = 35.50 -1.50 0.06
4 31 181(68)/362 = 34.00 -3.00 0.27
5 33 181(74)/362 = 37.00 -4.00 0.43
6 35 181(87)/362 = 43.50 -8.50 1.66
7 34 181(62)/362 = 31.00 3.00 0.29
8 34 181(71)/362 = 35.50 -1.50 0.06
9 37 181(68)/362 = 34.00 3.00 0.26
10 41 181(74)/362 = 37.00 4.00 0.43
Total 362 362 χ 2 = 5.41
between the days an employee is absent and the shift on which the employee works.
15.57 H 0 : The two variables (satisfaction and relationship) are independent

483
1 21 171(91)/447 = 34.81 -13.81 5.48
2 25 171(122)/447 = 46.67 -21.67 10.06
3 54 171(114)/447 = 43.61 10.39 2.48
4 71 171(120)/447 = 45.91 25.09 13.72
5 39 176(91)/447 = 35.83 3.17 0.28
6 49 176(122)/447 = 48.04 0.96 0.02
7 50 176(114)/447 = 44.89 5.11 0.58
8 38 176(120)/447 = 47.25 -9.25 1.81
9 31 100(91)/447 = 20.36 10.64 5 56
10 48 100(122)/447 = 27.29 20.71 15.71
11 10 100(114)/447 = 25.50 -15.50 9.42
12 11 100(120)/447 = 26.85 -15.85 9.35
Total 447 447 χ 2 = 74.47
χ 2 = 74.47, p-value = 0. There is sufficient evidence to infer that the level of job satisfaction
depends on boss/employee gender relationship.
15.58 H 0 : The two variables (Country and stress) are independent

A B C D E
1 Contingency Table
2
3 Stress
4 Country 1 2 TOTAL
5 1 266 315 581
6 2 347 276 623
7 3 153 187 340
8 4 164 128 292
9 5 92 79 171
10 TOTAL 1022 985 2007
11
12
14 df 4
15 p-value 0.0004
χ 2 = 20.3755, p-value = .0004. There is enough evidence to infer that Americans and Canadians
differ in their sources of stress.
484
15.59 H 0 : The two variables (method and quit) are independent
A B C D E F G
1 Contingency Table
2
3 Quit
4 Method 1 2 3 4 TOTAL
5 1 104 125 32 49 310
6 2 14 17 5 9 45
7 TOTAL 118 142 37 58 355
8
9
11 df 3
12 p-value 0.9009
χ 2 = .5803, p-value = .9009. There is not enough evidence to infer that the four methods differ in
their success rates.
15.60 H 0 : The two variables (education and section) are independent

A B C D E F G
1 Contingency Table
2
3 Section
4 Education 1 2 3 4 TOTAL
5 1 4 21 31 14 70
6 2 27 32 18 2 79
7 3 1 20 42 22 85
8 4 10 44 22 3 79
9 TOTAL 42 117 113 41 313
10
11
13 df 9
14 p-value 0
χ 2 = 86.6154, p-value = 0. There is sufficient evidence to infer that educational level affects the
way adults read the newspaper.
15.61a The expected frequency is 1/49.

b H 0 : p1 = 1/49, p 2 = 1/49, . . . , p 49 = 1/49
485
Number i f i ei (f i − e i ) (f i − ei ) 2 / ei
1 5 312(1/49) = 6.37 -1.38 0.29
2 6 312(1/49) = 6.37 -0.38 0.02
3 7 312(1/49) = 6.37 0.63 0.06
. .
. .
. .
47 6 312(1/49) = 6.37 -0.37 0.02
48 10 312(1/49) = 6.37 3.63 2.07
49 6 312(1/49) = 6.37 -0.37 0.02
Total 312 312 χ 2 = 38.22
χ 2 = 38.22, p-value = .8427. There is not enough evidence to infer that the numbers were not
generated randomly.
15.62 Binomial probabilities with n = 5 and p = .5: P(X = 0) = .0313, P(X = 1) = .1563, P(X = 2) =
.3125, P(X = 3) = .3125, P(X = 4) = .1563, P(X = 5) = .0313
H 0 : p 0 = .0313, p1 = .1563, p 2 = .3125, p 3 = .3125, p 4 = .1563, p 5 = .0313

0 8 200(.0313) = 6.26 1.74 0.48
1 35 200(.1563) = 31.26 3.74 0.45
2 57 200(.3125) = 62.50 -5.50 0.48
3 69 200(.3125) = 62.50 6.50 0.68
4 28 200(.1563) = 31.26 -3.26 0.34
5 3 200(.0313) = 6.26 -3.26 1.70
Total 200 200 χ 2 = 4.13
Rejection region: χ 2 > χ α2 ,6−1 = χ .205,5 = 11.1
χ 2 = 4.13, p-value = .5310. There is not enough evidence to infer that at the number of boys in
families with 5 children is not a binomial random variable with p =.5.
15.63 H 0 : The two variables (cold and group) are independent

486
A B C D E
1 Contingency Table
2
3 Group
4 Cold 1 2 TOTAL
5 1 17 11 28
6 2 12 13 25
7 3 9 18 27
8 4 16 18 34
9 TOTAL 54 60 114
10
11
13 df 3
14 p-value 0.2468
χ 2 = 4.139, p-value = .2468. There is not enough evidence to infer there are differences between
the four groups.
15.64 H 0 : The two variables (faculty and retire) are independent

A B C D E F G H
1 Contingency Table
2
3 Retire
4 Faculty 1 2 3 4 5 TOTAL
5 1 174 51 113 42 86 466
6 2 13 7 22 7 6 55
7 TOTAL 187 58 135 49 92 521
8
9
11 df 4
12 p-value 0.0452
χ 2 = 9.732, p-value = .0452. There is enough evidence to infer that whether a professor wishes to
retire is related to the faculty.
15.65 H 0 : The two variables (tree choice and age category) are independent
487
A B C D E F
1 Contingency Table
2
3 Tree choice
4 Age category 1 2 3 TOTAL
5 1 136 309 158 603
6 2 196 339 370 905
7 TOTAL 332 648 528 1508
8
9
11 df 2
12 p-value 0
χ 2 = 38.41, p-value = 0. There is enough evidence to infer that there are differences in the choice
of Christmas tree between the three age categories.
15.66 H 0 : The two variables (network and ask) are independent

A B C D E F
1 Contingency Table
2
3 Ask
4 Network 1 2 3 TOTAL
5 1 19 30 43 92
6 2 104 107 123 334
7 TOTAL 123 137 166 426
8
9
11 df 2
12 p-value 0.1016
χ 2 = 4.573, p-value = .1016. There is not enough evidence to conclude that there are differences
in responses between the three network news shows.
15.67 a H 0 : The two variables (education and group) are independent
488
A B C D E F G
1 Contingency Table
2
3 Group
5 1 5 113 73 40 231
6 2 70 305 189 55 619
7 TOTAL 75 418 262 95 850
8
9
11 df 3
12 p-value 0
χ 2 = 26.7059, p-value = 0. There is enough evidence to infer that there are differences in
educational attainment between those who belong and those who do not belong to the health
conscious group.
b H 0 : The two variables (education and buy Special X) are independent
A B C D E F G
1 Contingency Table
2
3 Buy Sp X
5 1 70 376 229 87 762
6 2 5 42 33 8 88
7 TOTAL 75 418 262 95 850
8
9
11 df 3
12 p-value 0.4007
between the four education groups and whether a person buys Special X.
15.68 a H 0 : The two variables (gender and vote) are independent
489
A B C D E
1 Contingency Table
2
3 Votes
4 Gender 1 2 TOTAL
5 1 189 169 358
6 2 203 204 407
7 TOTAL 392 373 765
8
9
11 df 1
12 p-value 0.4207
χ 2 = .6483, p-value = .4207. There is not enough evidence to infer that voting and gender are
related.
b H 0 : The two variables (education and vote) are independent
A B C D E F G
1 Contingency Table
2
3 Votes
4 Educ 1 2 3 4 TOTAL
5 1 48 164 107 39 358
6 2 34 178 134 61 407
7 TOTAL 82 342 241 100 765
8
9
11 df 3
12 p-value 0.0521
χ 2 = 7.7214, p-value = .0521. There is not enough evidence to infer that voting and educational
level are related.

c H 0 : The two variables (income category and vote) are independent
490
A B C D E F G
1 Contingency Table
2
3 Votes
4 Income 1 2 3 4 TOTAL
5 1 38 186 105 29 358
6 2 21 185 128 73 407
7 TOTAL 59 371 233 102 765
8
9
11 df 3
12 p-value 0
χ 2 = 23.108, p-value = 0. There is enough evidence to infer that voting and income are related.
15.69 H 0 : The two variables are (type of work and segment) independent
A B C D E F
1 Contingency Table
2
3 Segment
4 Work 1 2 3 TOTAL
5 1 157 44 217 418
6 2 219 53 264 536
7 3 256 102 524 882
8 TOTAL 632 199 1005 1836
9
10
12 df 4
13 p-value 0.0001
χ 2 = 23.0946, p-value = .0001. There is enough evidence to infer that there are differences in
employment between the three market segments.
15.70 H 0 : The two variables (value and segment) are independent

491
A B C D E F
1 Contingency Table
2
3 Segment
4 Value 1 2 3 TOTAL
5 1 147 135 136 418
6 2 221 155 160 536
7 3 339 254 289 882
8 TOTAL 707 544 585 1836
9
10
12 df 4
13 p-value 0.3411
χ 2 = 4.5122, p-value = .3411. There is not enough evidence to infer that there are differences in
the definition of value between the three market segments.
15.71 H 0 : The two variables (breakfast and group) are independent

A B C D E F G
1 Contingency Table
2
3 Group
4 Breakfast 1 2 3 4 TOTAL
5 1 3 15 29 222 269
6 2 44 53 95 292 484
7 3 17 10 79 135 241
8 4 25 77 77 77 256
9 TOTAL 89 155 280 726 1250
10
11
13 df 9
14 p-value 0
χ 2 = 206.4908, p-value = 0. There is enough evidence to infer that there are differences in
frequency of healthy breakfasts between the three four segments.
492
Appendix 15
A15.1 Chi-squared goodness-of-fit test

H 0 : p1 = .50 , p 2 = .20 , p 3 = .15 p 4 = .10 p 5 = .05
C D E
1 Actual Expected
2 183 175
3 63 70
4 55 52.5
5 29 35
6 20 17.5
7 p-value= 0.6321
p-value = .6321. There is not enough evidence to conclude that applicants to WLU's MBA
program are different in terms of their undergraduate degrees from the population of MBA
applicants?
A15.2 t-‐test of µ D
H 0 : µ D = 0
H1 : µ D < 0

A B C
2
3 First Sat Second SAT
4 Mean 1175 1190
5 Variance 28422 35392
9 df 39
10 t Stat -1.20
11 P(T<=t) one-tail 0.1182
13 P(T<=t) two-tail 0.2365
t = –1.20, p-value = .1182. There is not enough evidence to indicate that repeating the SAT
produces higher exam scores.
493

A15.3 Time to solve the 48 problems: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Diet Not
4 Mean 581.95 551.5
5 Variance 2716.6 2221.5
9 df 38
10 t Stat 1.94
11 P(T<=t) one-tail 0.0300
13 P(T<=t) two-tail 0.0601
t = 1.94, p-value = .0300. There is enough evidence to conclude that dieters take longer to solve
the 48 problems than do nondieters.
Successfully repeat string of five letters: z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) < 0
A B C D
2
3 Diet Not
7 z Stat -1.99
10 P(Z<=z) two-tail 0.0468
z = –1.99, p-value = .0234. There is enough evidence to conclude that dieters are less successful at
repeating string of five letters.
494

Successfully repeat string of five words: z-test of p1 − p 2 (case 1)
H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
A B C D
2
3 Diet Not
7 z Stat -1.58
10 P(Z<=z) two-tail 0.1134
z = –1.58, p-value = .0567. There is not enough evidence to conclude that dieters are less
successful at repeating string of five words.
A15.4 a t-estimator of µ

A B C D
1 t-Estimate: Mean
2
3 Overdue
4 Mean 7.09
6 LCL 6.40
7 UCL 7.77
LCL = 6.40, UCL = 7.77
b LCL = 50,000($.25)(6.40) = $80,000
UCL = 50,000($.25)(7.77) = $97,125
It does appear that not all fines are collected

H 0 : µ1 = µ 2 = µ 3
495

A B C D E F G
10 ANOVA
12 Between Groups 57512 2 28756 3.23 0.0468 3.16
13 Within Groups 506984 57 8894
14
15 Total 564496 59
F = 3.23; p-value = .0468. There is enough evidence to conclude that differences in sales exist
between the three advertising strategies.
A15.6 Chi-squared test of a contingency table

H 0 : The two variables (income category and mutual fund ownership) are independent
A B C D E
1 Contingency Table
2
3 Income category
4 Mutual fund 1 2 TOTAL
5 1 71 13 84
6 2 59 28 87
7 3 86 55 141
8 4 87 157 244
9 5 32 145 177
10 6 58 205 263
11 TOTAL 393 603 996
12
13
15 df 5
16 p-value 0
χ 2 = 196.77;p-value = 0. There is overwhelming evidence to infer that household income and
ownership of mutual funds are related
A15.7 Two-factor analysis of variance
496

A B C D E F G
23 ANOVA
25 Sample 13172 1 13172 1.42 0.2387 4.02
26 Columns 98839 2 49419 5.33 0.0077 3.17
27 Interaction 1610 2 805 0.09 0.9171 3.17
28 Within 501137 54 9280
29
30 Total 614757 59
Interaction: F = .09; p-value = .9171. There is no evidence of interaction.
Advertising strategy (Columns): F = 5.33; p-value = .0077. There is sufficient evidence to
conclude that advertising strategies differ with respect to sales.
Media (Sample): F = 1.42; p-value = .2387. There is not enough evidence to conclude that
differences in the medium for advertising differ in terms of sales.
A15.8 z-test of p1 − p 2 (case 1) Code 3 results were omitted.
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
A B C D
2
3 Folic acid Placebo
7 z Stat -2.85
10 P(Z<=z) two-tail 0.0044
z = –2.85, p-value = .0022. There is overwhelming evidence to conclude that folic acid reduces the
incidence of spina bifida.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
497

A B C
2
3 British American
4 Mean 238.0 252.0
5 Variance 149.9 220.2
8 df 531
9 t Stat -12.00
10 P(T<=t) one-tail 7.64E-30
12 P(T<=t) two-tail 1.53E-29
z = –12.00, p-value = 0. There is enough evidence to infer that British golfers play golf in less
time than do American golfers.
A15.10 one-way analysis of variance

H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 626046 2 313023 58.37 0.0000 3.03
13 Within Groups 1523047 284 5363
14
15 Total 2149093 286
F = 58.37, p-value = 0. There is enough evidence to conclude that there are differences between
the three groups.
Multiple Comparisons
A B C D E
2
3 LSD Omega
5 Before 1976 After 1986 122.62 28.03 25.46
6 Canadian 78.04 24.67 25.46
7 After 1986 Canadian -44.58 25.75 25.46
All three groups differ from each other.
498

H 0 : The two variables (year and sport) are independent
A B C D E
1 Contingency Table
2
3 Sport
4 Year 1 2 TOTAL
5 1 116 122 238
6 2 119 92 211
7 3 29 58 87
8 4 52 39 91
9 5 48 34 82
10 6 16 33 49
11 7 26 29 55
12 8 24 21 45
13 9 70 72 142
14 TOTAL 500 500 1000
15
16
18 df 8
19 p-value 0.0025
χ 2 = 23.8101, p-value = .0025. There is enough evidence to infer that North Americans changed
their favorite sport between 1985 and 1992.
A B C D
1 t-Estimate: Mean
2
3 Cars
4 Mean 165.79
6 LCL 157.17
7 UCL 174.41
Five minute interval: LCL = 157.17, UCL = 174.41

Twenty-four hour day (12 5-minute intervals, 24 hours per day):
LCL = 12 × 24 × 157.17 = 45,265
UCL = 12 × 24 × 174.41 = 50,230
499

A15.13 z-estimator of p
A B
2 Exercise?
4 Observations 671
5 LCL 0.514
6 UCL 0.589
Total number of adults who exercise:
LCL = 205.9 million (.514) = 105.8 million
UCL = 205.9 million (.589) = 121.3 million
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Activity Usual
4 Mean 57.06 87.28
5 Variance 296.18 215.42
8 Hypothesized Mean Difference 0.00
9 df 132
10 t Stat -10.94
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = –10.94, p-value = 0. There is enough evidence to indicate to infer that graded activity is
effective.

H 0 : The two variables (group and improvement) are independent
500

A B C D E
1 Contingency Table
2
3 Group
4 Improvement 1 2 TOTAL
5 1 42 8 50
6 2 32 18 50
7 3 13 37 50
8 TOTAL 87 63 150
9
10
12 df 2
13 p-value 0
χ 2 = 35.63, p-value = 0. There is sufficient evidence to infer there are differences between the
three groups.
A B C D E
2
4 Sample size 780 0.774 ± 0.0294
Total number of on-time departures:

LCL = 7,140,596(.7446) = 5,316,888
UCL = 7,140,596(.8034) = 5,736,755
A15.17 H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Two-tail F test: F = 1.00, p-value .9902 Use equal variances t-test.
501

t = 3.10, p-value = .0019. There is enough evidence to infer that job tenure changed between 2006
and 2008.

H 0 : The two variables (party and support for capital punishment) are independent
χ 2 = 110.35, p-value = 0. There is enough evidence to conclude that Democrats, Republicans, and
Independents differ in their support for capital punishment.
A15.19 t-estimate of a mean
502

LCL = 23.49, UCL = 24.04

H 0 : The two variables (PARTYID and SEX) are independent
χ 2 = 41.27, p-value = 0. There is sufficient evidence to conclude that men and women differ in
their political affiliation.

H 0 : µ1 = µ 2 = µ 3 = µ 4
503

F = 21.86, p-value = 6.31E-14 ≈ 0. There is enough evidence to conclude that there are differences
between classes with respect to the number of family earners.

H 0 : µ1 = µ 2 = µ 3 = µ 4
F = 14.45, p-value = 5.5E-07 ≈ 0. There is enough evidence to infer that inflation-adjusted

incomes differ between 2002, 2004, and 2006.
504

H 0 : The two variables (RACE and WRKSLF) are independent
χ 2 = 13.87, p-value = .0010. There is sufficient evidence to conclude that differences exist
between the races in whether an individual is self-employed.

H 0 : The two variables (PARTYID3 and UNEMP) are independent
505

χ 2 = 14.61, p-value = .0007. There is sufficient evidence to conclude that Americans who have
been unemployed in the last ten years have different party affiliations.
A15.25 H0: (µ1 - µ2) = 0

H1: (µ1 - µ2) ≠ 0
Two-tail F test: F = .866 p-value = .1339. Use equal-variances
t =.600, p-value = .5485. There is not enough evidence to infer that Liberals and Conservatives
differ in their intention to vote.

H 0 : The two variables (Year and EMPLOY) are independent
506

χ 2 = 2077, p-value = 0. There is sufficient evidence to conclude that employment status changed
between 2004 and 2008.
A15.27 H0: (µ1 - µ2) = 0

H1: (µ1 - µ2) > 0
Two-tail F test: F = 1.14, p-value = .1429. Use equal-variances t-test
t = .152, p-value = .8795. There is not sufficient evidence to conclude that American who consider
themselves strong Democrats or Republicans have more education than those who are not strong
Democrats or Republicans.
Case A15.1 One-way analysis of variance

H 0 : µ1 = µ 2 = µ 3 = µ 4
Weight loss:
A B C D E F G
11 ANOVA
13 Between Groups 189.4 3 63.14 1.78 0.1532 2.66
14 Within Groups 5533.6 156 35.47
15
16 Total 5723.0 159
weight loss between the four diets.
507

Percent LDL decrease:
A B C D E F G
11 ANOVA
13 Between Groups 2569.4 3 856.48 32.56 0.0000 2.66
14 Within Groups 4104.0 156 26.31
15
16 Total 6673.4 159
F = 32.56, p-value = 0. There is enough evidence to conclude that there are differences in bad
cholesterol reduction between the four diets.
Percent HDL Increase:

A B C D E F G
11 ANOVA
13 Between Groups 5595.0 3 1864.99 92.62 0.0000 2.66
14 Within Groups 3141.3 156 20.14
15
16 Total 8736.2 159
F = 96.62, p-value = 0. There is enough evidence to conclude that there are differences in good
cholesterol increase between the four diets.
508

Chapter 16
16.1 a The slope coefficient tells us that for additional inch of father’s height the son’s height
increases on average by .516. The y-intercept is meaningless.
b On average the son will be shorter than his father.
c On average the son will be taller than his father.
16.2 a
Scatter Diagram
20
15
Sales
10
0
0 20 40 60 80 100 120
Advertising
b xi yi x i2 yi2 x i yi
23 9.6 529 92.16 220.8
46 11.3 2,116 127.69 519.8
60 12.8 3,600 163.84 768.0
54 9.8 2,916 96.04 529.2
28 8.9 784 79.21 249.2
33 12.5 1,089 156.25 412.5
25 12.0 625 144.00 300.0
31 11.4 961 129.96 353.4
36 12.6 1,296 158.76 453.6
88 13.7 7,744 187.69 1205.6
90 14.4 8,100 207.36 1296.0
99 15.9 9,801 252.81 1,574.1
Total 613 144.9 39,561 1,795.77 7,882.2
n n n n
∑ x i = 613 ∑ y i = 144.9 ∑ x i2 = 39,561 ∑x y i i = 7,882.2

i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑y
i =1
xi
i =1
i ⎥
⎥ 1 ⎡ (613)(144.9 ⎤
s xy = ∑
n − 1 ⎢⎢ i =1
x i yi −
n ⎥ = 12 − 1 ⎢7,882.2 −
⎣ 12 ⎥ = 43.66
⎦
⎥
⎢ ⎥
⎣ ⎦
509
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎥
⎟
1 ⎢ n ⎜ ∑
⎟ ⎥
1 ⎡ (613) 2 ⎤
s 2x = ⎢ x i2 − ⎝ i =1 ⎠ ⎥ =
∑ ⎢ 39,561 − ⎥ = 749.7
n − 1 ⎢ i =1 n ⎥ 12 − 1 ⎣⎢ 12 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 43.66
b1 = = = .0582
s 2x 749.7
x=
∑x i
=
613
= 51.08
n 12
y=
∑y i
=
144.9
= 12.08
n 12
b 0 = y − b1x = 12.08 – (.0582)(51.08) = 9.107
The sample regression line is
ŷ = 9.107 + .0582x
The slope tells us that for each additional thousand dollars of advertising sales increase on average
by .0582 million. The y-intercept has no practical meaning.
16.3 a xi yi x i2 yi2 x i yi
8.5 115 72.25 13,225 977.5
7.8 111 60.84 12,321 865.8
7.6 185 57.76 34,225 1,406.0
7.5 201 56.25 40,401 1,507.5
8.0 206 64.00 42,436 1,648.0
8.4 167 70.56 27,889 1,402.8
8.8 155 77.44 24,025 1,364.0
8.9 117 79.21 13,689 1,041.3
8.5 133 72.25 17,689 1,130.5
8.0 150 64.00 22,500 1,200.0
Total 82.0 1,540 674.56 248,400 12,543.4
n n n n
∑ x i = 82.0 ∑ y i = 1,540 ∑ x i2 = 674.56 ∑x y i i = 12,543.4

i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑y
i =1
xi
i =1
i ⎥
⎥ 1 ⎡ (82.0)(1,540 ⎤
s xy =
n − 1 ⎢⎢ i =1∑x i yi −
n ⎥ = 10 − 1 ⎢12,543.4 −
⎣ 10 ⎥ = −9.40
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(82.0) 2 ⎤
⎢ x i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢674.56 −
1 ⎡
s 2x = ∑ ⎥ = .24
n − 1 ⎢ i =1 n ⎥ 10 − 1 ⎣⎢ 10 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
510
s xy −9.40
b1 = = = −39.17
s 2x .24
x=
∑x i
=
82.0
= 8.20
n 10
y=
∑y i
=
1,540
= 154.0
n 10
b 0 = y − b1x = 154.0 – (–39.17)(8.20) = 475.2
ŷ = 475.2 – 39.17x
b. The slope coefficient tells us that for each additional 1 percentage point increase in mortgage
rates, the number of housing starts decreases on average by 39.17. The y-intercept has no
meaning.
16.4a
Scatter Diagram
20
15
10
Overweight
0
0 10 20 30 40 50
-‐5
-‐10
-‐15
Television
42 18 1,764 324 756
34 6 1,156 36 204
25 0 625 0 0
35 –1 1,225 1 –35
37 13 1,369 169 481
38 14 1,444 196 532
31 7 961 49 217
33 7 1,089 49 231
19 –9 361 81 –171
29 8 841 64 232
38 8 1,444 64 304
28 5 784 25 140
511
29 3 841 9 87
36 14 1,296 196 504
18 –7 324 49 –126
Total 472 86 15,524 1,312 3,356
n n n n
∑ x i = 472 ∑ y i = 86 ∑ x i2 = 15,524 ∑x y i i = 3,356

i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑y
i =1
xi
i =1
i ⎥
⎥ 1 ⎡ (472)(86 ⎤
s xy =
n − 1 ⎢⎢ i =1∑x i yi −
n ⎥ = 15 − 1 ⎢3,356 − 15 ⎥ = 46.42
⎣ ⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(472) 2 ⎤
⎢ x i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢15,524 −
1 ⎡
s 2x = ∑ ⎥ = 47.98
n − 1 ⎢ i =1 n ⎥ 15 − 1 ⎣⎢ 15 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 46.42
b1 = = = .9675
s 2x 47.98
x=
∑x i
=
472
= 31.47
n 15
y=
∑y i
=
86
= 5.73
n 15
b 0 = y − b1x = 5.73 – (.9675)(31.47) = –24.72
ŷ = –24.72 + .9675x
The slope coefficient indicates that for each additional hour of television weight increases on
average by .9675 pounds. The y-intercept is the point at which the regression line hits the y–axis;
it has no practical meaning.
16.5a xi yi x i2 yi2 x i yi
80 20,533 6,400 421,604,089 1,642,640
68 1,439 4,624 2,070,721 97,852
78 13,829 6,084 191,241,241 1,078,662
79 21,286 6,241 453,093,796 1,681,594
87 30,985 7,569 960,070,225 2,695,695
74 17,187 5,476 295,392,969 1,271,838
86 30,240 7,396 914,457,600 2,600,640
92 37,596 8,464 413,459,100 3,458,832
77 9,610 5,929 92,352,100 739,970
84 28,742 7,056 826,102,564 2,414,328
Total 805 211,447 65,239 5,569,844,521 17,682,051
512
n n n n
2
∑x
i =1
i = 805 ∑y
i =1
i = 211,447 ∑x
i =1
i = 65,239 ∑x y
i =1
i i = 17,682,051
⎡ n n ⎤
1 ⎢
⎢ n ∑x ∑y
i =1
i
i =1
i ⎥
⎥ 1 ⎡ (805)(211,447 ⎤
s xy =
n − 1 ⎢⎢ i =1 ∑
x i yi −
n ⎥ = 10 − 1 ⎢17,682,051 −
⎣ 10 ⎥ = 73,396
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
1 ⎢ n ⎜
⎝ i 1
⎟ ⎥∑
⎠ ⎥ = 1 ⎡ (805) 2 ⎤
s 2x = ⎢ x i2 − =
n − 1 ⎢ i =1 ∑ n
⎢65,239 −
⎥ 10 − 1 ⎢⎣
⎥ = 48.50
10 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 73,396
b1 = = = 1,513
s 2x 48.50
x=
∑x i
=
805
= 80.5
n 10
y=
∑y i
=
211,447
= 21,145
n 10
b 0 = y − b1x = 21,145 – (1,513)(80.5) = –100,652
ŷ = –100,652 + 1,513x
b. For each additional one degree increase in temperature the number of beers sold increases on
average by 1,513. The y-intercept is the point at which the regression line hits the y–axis; it has no
practical meaning.
513
16.6 a
Scatter Diagram
30
25
20
Test scores
15
10
5
0
0 20 40 60 80
Lengths
s xy 51.86
b b1 = = = .2675, b 0 = y − b1x = 13.80 – .2675(38.00) = 3.635
s 2x 193.9
Regression line: ŷ = 3.635 + .2675x (Excel: ŷ = 3.636 + .2675x)
c b1 = .2675; for each additional second of commercial, the memory test score increases on
average by .2675. b0 = 3.64 is the y-intercept.
s xy 86.93
16.7a b1 = = = 1.465, b 0 = y − b1x = 210.4 – 1.465(13.68) =190.4.
s 2x 59.32
Regression line: ŷ = 190.4 + 1.465x (Excel: ŷ = 190.4 + 1.465x)
b For each additional floor prices increase on average by $1.465 thousand ($1,465). The y-
intercept has no practical meaning.
s xy 9.67
16.8 b1 = = = .0899, b 0 = y − b1x = 11.55 – .0899(45.49) =7.460.
s 2x 107.51
Regression line: ŷ =7.460 + .0899x (Excel: ŷ = 7.462 + .0900x)
The slope coefficient tells us that for each additional year of age time increases on average by
.0899 minutes. The y-intercept has no meaning.
s 2xy −6.44
16.9 a b 1 = = = −.1169, b 0 = y − b1x = 26.28 – (–.11697)(37.29) = 30.64.
s 2x 55.11
514
Regression line: ŷ = 30.64 – .1169x (Excel: ŷ = 30.63 – .1169x)
b The slope coefficient indicates that for each additional year of age, the employment period
decreases on average by .1169. b0 = 30.63 is the y-intercept.
s xy 20.55
16.10a b1 = = = .1898, b 0 = y − b1x = 14.43 – .1898(37.64) = 7.286.
s 2x 108.3
Regression line: ŷ = 7.286 +.1898x (Excel: ŷ = 7.287 +.1897x)
b For each additional cigarette the number of days absent from work increases on average by
.1898. The y-intercept has no meaning.
s xy 22.83
16.11 b1 = = = 5.347, b 0 = y − b1x = 49.22 – 5.347(4.885) = 23.10.
s 2x 4.270
For each addition kilometer a house is away from its nearest fire station the percentage damage
increases on average by 5.347.
s xy 30,945
16.12a b1 = = = 44.97, b 0 = y − b1x = 6,465 – 44.97(53.93) = 4040.
s 2x 688.2
Regression line: ŷ = 4040 + 44.97x (Excel: ŷ = 4040 + 44.97x)
b. For each additional thousand square feet the price increases on average by $44.97 thousand.
s xy −81.78
16.13 b1 = = = −.00138, b 0 = y − b1 x = 27.73 – (–.00138)(1199) = 29.39.
s 2x 59,153
Regression line: ŷ = 29.39–.00138x (Excel: 29.39–.00138x)
For each additional hour the price decreases on average by .00138 thousand dollars or $1.38.
s xy 310.0
16.14 b1 = = = 64.05, b 0 = y − b1x = 762.6 –64.05(4.75) = 458.4.
s 2x 4.84
For each additional occupant the electrical use increases on average by 64.05.
s xy 225.7
16.15 b1 = = = 1.959, b 0 = y − b1x = 270.3 –1.959(59.42) = 153.9.
s 2x 115.2
Regression line: ŷ = 153.9 + 1.959x (Excel: : ŷ = 153.9 + 1.958x)
For each additional $1,000 of income the weekly food budget increases on average by $1.96.
515
s xy −10.78
16.16 a b1 = = = −.3039, b 0 = y − b1x = 17.20 – (–.3039)(11.33) = 20.64.
s 2x 35.47
Regression line: ŷ = 20.64 – .3039x (Excel: ŷ = 20.64 – .3038x)
b The slope indicates that for each additional one percentage point increase in the vacancy rate
rents on average decrease by $.3039.
s xy 6.020
16.17 a b1 = = = .604 , b 0 = y − b1x = 59.59 –.604(68.95) =17.94.
s 2x 9.966
b For each additional inch of height income increases on average by $.604 thousand or $604.
s xy .8258
16.18 b1 = = = .0514, b 0 = y − b1x = 93.89 –.0514(79.47) = 89.81.
s 2x 16.07
For each additional mark on the test the number of non-defective products increases on average by
.0514.
16.19 For each commercial length, the memory test scores are normally distributed with constant
variance and a mean that is a linear function of the commercial lengths.
16.20 For each number of years of education incomes are normally distributed with constant
variance and a mean that is a linear function of the number of years of education.
16.21 For each number of hours prices are normally distributed with constant variance and a mean
that is a linear function of the number of hour.
16.22 b x i yi x i2 yi2 x i yi
1 1 1 1 1
3 8 9 64 24
4 15 16 225 60
6 33 36 1089 198
9 75 81 5625 675
8 70 64 4900 560
10 95 100 9025 950
Total 41 297 307 20,929 2,468
n n n n n
∑ x i = 41 ∑ y i = 297 ∑ x i2 = 307 ∑ y i2 = 20,929 ∑x y i i = 2,468

i =1 i =1 i =1 i =1 i =1
516
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑
i =1
xi
i =1
y i ⎥
⎥ 1 ⎡ (41)(297 ⎤
s xy =
n − 1 ⎢⎢ i =1∑x i yi −
n ⎥ = 7 − 1 ⎢2,468 −
⎣ 7 ⎥ = 121.4
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
1 ⎢ n ⎜
⎝ i 1
∑
⎟ ⎥
⎠ ⎥ = 1 ⎡ (41) 2 ⎤
s 2x = ⎢ x i2 − =
n − 1 ⎢ i =1∑ n
⎢307 −
⎥ 7 − 1 ⎢⎣
⎥ = 11.14
7 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(297) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢20,929 −
1 ⎡
s 2y = ∑ ⎥ = 1,388.0
n − 1 ⎢ i =1 n ⎥ 7 − 1 ⎣⎢ 7 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 121.4
b1 = = = 10.90
s 2x 11.14
⎛ 2 s 2xy ⎞ ⎛ (121.4) 2 ⎞
SSE = (n − 1)⎜ s y − 2 ⎟ = (7 − 1)⎜1,388.0 − ⎟ = 390.1
⎜ s x ⎟⎠ ⎜ 11.14 ⎟
⎝ ⎝ ⎠
SSE 390.1
sε = = = 8.83 (Excel: s ε = 8.85)
n−2 7−2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,5 = 2.571 or t < − t α / 2,n − 2 = − t .025,5 = −2.571
sε 8.83
s b1 = = = 1.08
(n − 1)s 2x (7 − 1)(11.14)
b 1 − β1 10.90 − 0
t= = = 10.09 (Excel: t = 10.07, p–value = .0002. There is enough evidence to
s b1 1.08
infer a linear relationship.
517
Scatter Diagram
120
100
80
60
40
20
0
0 2 4 6 8 10 12
-‐20
There does appear to be a linear relationship.
16.23a
Scatter Diagram
300
250
200
150
100
50
0
0 1 2 3 4 5 6 7
-‐50
There does appear to be a linear relationship.
3 25 9 625 75
5 110 25 12100 550
2 9 4 81 18
6 250 36 62500 1500
1 3 1 9 3
4 71 16 5041 284
Total 21 468 91 80,356 2,430
518
n n n n n
2 2
∑x
i =1
i = 21 ∑y
i =1
i = 468 ∑x
i =1
i = 91 ∑y
i =1
i = 80,356 ∑x y
i =1
i i = 2,430
⎡ n n ⎤
1 ⎢
⎢ n ∑x ∑y
i =1
i
i =1
i ⎥
⎥ 1 ⎡ (21)(468) ⎤
s xy =
n − 1 ⎢⎢ i =1 ∑
x i yi −
n ⎥ = 6 − 1 ⎢2,430 −
⎣ 6 ⎥ = 158.4
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎟ ⎥
1 ⎢ n ⎜
⎝ i 1
⎟ ⎥ ∑
⎠ ⎥ = 1 ⎡ (21) 2 ⎤
s 2x = ⎢ x i2 − =
n − 1 ⎢ i =1 ∑ n
⎢91 −
⎥ 6 − 1 ⎢⎣
⎥ = 3.50
6 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎟ ⎥
1 ⎢
n ⎜ ⎟ ⎥ ∑ (468) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢80,356 −
⎡
s 2y = ∑ ⎥ = 8,770
n − 1 ⎢ i =1 n ⎥ 6 − 1 ⎢⎣ 6 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy 158.4
b1 = = = 45.26
s 2x 3.5
⎛ s 2xy ⎞ ⎛ (158.4) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (6 − 1)⎜ 8,770 − ⎟ = 8,006
⎜ s x ⎟⎠ ⎜ 3.50 ⎟
⎝ ⎝ ⎠
SSE 8006
sε = = = 44.74 (Excel: s ε = 44.75)
n−2 6−2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t .025, 4 = 2.776 or t < − t α / 2,n − 2 = − t .025, 4 = −2.776
sε 44.74
s b1 = = = 10.69
(n − 1)s 2x (6 − 1)(3.50)
b 1 − β1 45.26 − 0
s b1 10.69
infer a linear relationship.
519
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(144.9) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢1,795.77 −
1 ⎡
16.24 a s 2y = ∑ ⎥ = 4.191
n − 1 ⎢ i =1 n ⎥ 12 − 1 ⎣⎢ 12 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎛ s 2xy ⎞ ⎛ (43.66) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (12 − 1)⎜ 4.191 − ⎟ = 18.13
⎜ s x ⎟⎠ ⎜ 749.7 ⎟
⎝ ⎝ ⎠
SSE 18.13
sε = = = 1.347 (Excel: s ε = 1.347)
n−2 12 − 2
b H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t.025,10 = 2.228 or t < − t α / 2,n − 2 = − t.025,10 = −2.228
sε 1.347
s b1 = = = .0148
(n − 1)s 2x (12 − 1)(749.7)
b 1 − β1 .0582 − 0
s b1 .0148
infer a linear relationship between advertising and sales.

c b1 ± t α / 2,n −2 s b = .0582 ± 2.228(.0148) = .0582 ± .0330 LCL = .0252, UCL = .0912
1
s 2xy (43.66) 2
d R2 = = = .6067 (Excel: R 2 = .6066). 60.67% of the variation in sales is
s 2x s 2y (749.7)(4.191)
explained by the variation in advertising.

e There is evidence of a linear relationship. For each additional dollar of advertising sales increase,
on average by .0582.
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎟ ⎥
1 ⎢
n ⎜ ∑
⎟ ⎥
(1,540) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢248,400 −
⎡
16.25 s 2y = ∑ ⎥ = 1249
n − 1 ⎢ i =1 n ⎥ 10 − 1 ⎢⎣ 10 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
s 2xy (−9.40) 2
R2 = = = .2948 (Excel: R 2 = .2948).
s 2x s 2y (.24)(1,249)
⎛ s 2xy ⎞ ⎛ (−9.49) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (10 − 1)⎜1,249 − ⎟ = 7,864
⎜ s x ⎟⎠ ⎜ .24 ⎟
⎝ ⎝ ⎠
520
SSE 7,864
sε = = = 31.35 (Excel: s ε = 31.48)
n−2 10 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t .025,8 = 2.306 or t < − t α / 2,n − 2 = − t .025,8 = −2.306
sε 31.35
s b1 = = = 21.33
(n − 1)s 2x (10 − 1)(.24)
b 1 − β1 −39.17 − 0
t= = = −1.84 (Excel: t = 1.83, p–value = .1048. There is not enough evidence
s b1 .21.33
to infer a linear relationship between interest rates and housing starts.
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑ ⎟ ⎥
(86) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢1,312 −
1 ⎡
16.26 s 2y = ∑ ⎥ = 58.50
n − 1 ⎢ i =1 n ⎥ 15 − 1 ⎣⎢ 15 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎛ s 2xy ⎞ ⎛ (46.42) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (15 − 1)⎜ 58.50 − ⎟ = 190.2
⎜ s x ⎟⎠ ⎜ 47.98 ⎟
⎝ ⎝ ⎠
SSE 190.2
sε = = = 3.825
n−2 15 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,13 = 2.160 or t < − t α / 2,n − 2 = − t .025,13 = −2.160
sε 3.825
s b1 = = = .1476
(n − 1)s 2x (15 − 1)(47.98)
b 1 − β1 .9675 − 0
t= = = 6.55 (Excel: t = 6.55, p–value = 0.) There is enough evidence to
s b1 .1476
conclude that there is a linear relationship between hours of television viewing and how
overweight the child is.
521
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑⎟ ⎥
(211,447) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢5,569,844,521 −
1 ⎡
16.27 s 2y = ∑ ⎥ = 122,095,682
n − 1 ⎢ i =1 n ⎥ 10 − 1 ⎣⎢ 10 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎛ s 2xy ⎞ ⎛ (73,396) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (10 − 1)⎜122,095,682 − ⎟ = 99,216,698
⎜ s x ⎟⎠ ⎜ 48.50 ⎟
⎝ ⎝ ⎠
SSE 99,216,698
sε = = = 3,522
n−2 10 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t .025,8 = 2.306 or t < − t α / 2,n − 2 = − t .025,8 = −2.306
sε 3522
s b1 = = = 168.6
(n − 1)s 2x (10 − 1)(48.50)
b 1 − β1 1,513 − 0
t= = = 8.97 (Excel: t = 8.98, p–value = 0.) There is evidence of a linear
s b1 168.6
relationship between temperature and the number of beers sold.
⎛ s 2xy ⎞ ⎛ (51.86) 2 ⎞
16.28 a SSE = (n − 1)⎜ s 2y − 2 ⎟ = (60 − 1)⎜ 47.96 − ⎟ = 2,011
⎜ s x ⎟⎠ ⎜ 193.9 ⎟
⎝ ⎝ ⎠
SSE 2,011
sε = = = 5.888 (Excel: s ε = 5.888). Relative to the values of the dependent
n−2 60 − 2
variable the standard error of estimate appears to be large indicating a weak linear relationship.
s 2xy (51.86) 2
b R2 = = = .2892 (Excel: R 2 = .2893).
s 2x s 2y (193.9)(47.96)
c H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,58 ≈ 2.000 or t < − t α / 2,n − 2 = − t .025,58 = −2.000
sε 5.888
s b1 = = = .0550
(n − 1)s 2x (60 − 1)(193.9)
b 1 − β1 .2675 − 0
t= = = 4.86 (Excel: t = 4.86, p–value = 0). There is enough evidence to infer a
s b1 .0550
linear relationship between memory test scores and length of commercial.
522
d b1 ± t α / 2,n −2 s b = .2675 ± 1.671(.0550) = .2675 ± .0919 LCL = .1756, UCL = .3594
1
⎛ s 2xy ⎞ ⎛ (86.93) 2 ⎞
16.29 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (50 − 1)⎜ 496.4 − ⎟ = 18,081
⎜ s x ⎟⎠ ⎜ 59.32 ⎟
⎝ ⎝ ⎠
SSE 18,081
n−2 50 − 2
s 2xy (86.93) 2
R2 = = = .2566 (Excel: R 2 = .2566).
s 2x s 2y (59.32)(496.4)
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025, 48 ≈ 2.009 or t < − t α / 2,n − 2 = − t .025, 48 = −2.009
sε 19.41
s b1 = = = .3600
(n − 1)s 2x (50 − 1)(59.32)
b 1 − β1 1.465 − 0
t= = = 4.07 (Excel: t = 4.07, p–value = .0002). There is evidence of a linear
s b1 .3600
relationship. The relationship however, is weak.
⎛ s 2xy ⎞ ⎛ (9.67) 2 ⎞
16.30 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (229 − 1)⎜ 42.54 − ⎟ = 9500.8
⎜ s ⎟ ⎜ 107 .51 ⎟
⎝ x ⎠ ⎝ ⎠
SSE 9500.8
sε = = = 6.47
n−2 229 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n −2 = t .025,227 ≈ 1.96 or t < − t α / 2,n −2 = − t .025,227 = −1.96
sε 6.47
s b1 = = = .0413
(n − 1)s 2x (229 − 1)(107.51)
b 1 − β1 .0899 − 0
t= = = 2.17 (Excel: t =2.18, p–value = .0305.) There is evidence of a linear
s b1 .0413
relationship between age and time to complete census.
⎛ s 2xy ⎞ ⎛ (−6.44) 2 ⎞
16.31 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (80 − 1)⎜ 4.00 − ⎟ = 256.5
⎜ s x ⎟⎠ ⎜ 55.11 ⎟
⎝ ⎝ ⎠
523
SSE 256.5
n−2 80 − 2
s 2xy (−6.44) 2
R2 = = = .1881 (Excel: R 2 = .1884). There is a weak linear relationship
s 2x s 2y (55.11)(4.00)
between age and number of weeks of employment.
⎛ s 2xy ⎞ ⎛ (20.55) 2 ⎞
16.32 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (231 − 1)⎜19.80 − ⎟ = 3657
⎜ s x ⎟⎠ ⎜ 108.3 ⎟
⎝ ⎝ ⎠
SSE 3657
sε = = = 3.996
n−2 231 − 2
H 0 : β1 = 0
H1 : β1 > 0
Rejection region: t > t α / 2,n − 2 = t .025, 229 ≈ 1.960 or t < − t α / 2,n − 2 = − t .025, 229 ≈ −1.960
sε 3.996
s b1 = = = .02532
(n − 1)s 2x (231 − 1)(108.3)
b 1 − β1 .1898 − 0
t= = = 7.50 (Excel: t =7.49, p–value = 0.) There is evidence of a positive linear
s b1 .02532
relationship between cigarettes smoked and the number of sick days.
⎛ s 2xy ⎞ ⎛ (22.83) 2 ⎞
16.33 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (85 − 1)⎜ 243.9 − ⎟ = 10,234
⎜ s x ⎟⎠ ⎜ 4.270 ⎟
⎝ ⎝ ⎠
SSE 10,234
sε = = = 11.10 (Excel: s ε = 11.11).
n−2 85 − 2
a H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,83 ≈ 1.990 or t < − t α / 2,n − 2 = − t .025,83 ≈ −1.990
sε 11.10
s b1 = = = .5861
(n − 1)s 2x (85 − 1)(4.270)
b 1 − β1 5.347 − 0
t= = = 9.12 (Excel: t =9.12, p–value = 0.) There is evidence of a linear
s b1 .5861
relationship between distance and fire damage.

b b1 ± t α / 2,n −2 s b = 5.347 ± 1.988(.5861) = 5.347 ± 1.166 LCL = 4.18, UCL = 6.51
1
524
s 2xy ( 22.83) 2
c R2 = = = .5005 (Excel: R 2 = .5004). There is a moderately strong linear
s 2x s 2y (4.270)(243.9)
relationship between distance and fire damage.
⎛ s 2xy ⎞ ⎛ (30,945) 2 ⎞
16.34 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (40 − 1)⎜11,918,489 − ⎟ = 410,554,683
⎜ s x ⎟⎠ ⎜ 688.2 ⎟
⎝ ⎝ ⎠
SSE 410,554,683
a sε = = = 3,287 (Excel: s ε = 3,287). There is a weak linear relationship.
n−2 40 − 2
b H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,38 ≈ 2.021or t < − t α / 2,n − 2 = − t .025,38 ≈ −2.021
sε 3,287
s b1 = = = 20.06
(n − 1)s 2x (40 − 1)(688.2)
b 1 − β1 44.97 − 0
t= = = 2.24 (Excel: t = 2.24, p–value = .0309.) There is enough evidence of a
s b1 20.06
s 2xy (30,945) 2
c R2 = = = .1167 (Excel: R 2 = .1168) 11.67% of the variation in
s 2x s 2y (688.2)(11,918,489)
percent damage is explained by the variation in distance to the fire station.
⎛ s 2xy ⎞ ⎛ (−81.78) 2 ⎞
16.35 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (60 − 1)⎜ 3.623 − ⎟ = 207.1
⎜ s x ⎟⎠ ⎜ 59,153 ⎟
⎝ ⎝ ⎠
SSE 207.1
sε = = = 1.890 (Excel: s ε =1.889 ).
n−2 60 − 2
H 0 : β1 = 0
H1 : β1 < 0
Rejection region: t < − t α,n − 2 = − t .05,58 ≈ −1.671
sε 1.890
s b1 = = = .001012
(n − 1)s 2x (60 − 1)(59,153)
b 1 − β1 −.00138 − 0
t= = = −1.364 (Excel: t = –1.367, p–value = .1769/2 = .0885.) There is not
s b1 .001012
enough evidence to infer that as hours of engine use increase the price decreases.
525
⎛ s 2xy ⎞
⎟ = (200 − 1)⎛⎜ 56,725 − (310.0)
2 ⎞
16.36 SSE = (n − 1)⎜ s 2y − ⎟ = 7,337,056
⎜
⎝ s 2x ⎟
⎠
⎜
⎝ 4.84 ⎟
⎠
SSE 7,337,056
sε = = = 191.1(Excel: s ε = 192.5).
n−2 200 − 2
s 2xy (310.0) 2
R2 = = = .3500 (Excel: R 2 = .3496) 35.00% of the variation in the
s 2x s 2y (4.84)(56,725)
electricity use is explained by the variation in the number of occupants.

H 0 : β1 = 0
H1 : β1 ≠ 0
sε 191.1
s b1 = = = 6.16
(n − 1)s 2x (200 − 1)(4.84)
b 1 − β1 64.05 − 0
t= = = 10.39 (Excel: t =10.32, p–value = 0.) There is enough evidence of a
s b1 6.16
s 2xy (225.7) 2
16.37 a R 2 = = = .2461 (Excel: R 2 = .2459) 24.61% of the variation in food
s 2x s 2y (115.2)(1,797)
budgets is explained by the variation in household income.
⎛ s 2xy ⎞ ⎛ (225.7) 2 ⎞
b SSE = (n − 1)⎜ s 2y − 2 ⎟ = (150 − 1)⎜1,797 − ⎟ = 201,866
⎜ s x ⎟⎠ ⎜ 115.2 ⎟
⎝ ⎝ ⎠
SSE 201,866
sε = = = 36.93 (Excel: s ε = 36.94 ).
n−2 150 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
sε 36.93
s b1 = = = .2819
(n − 1)s 2x (150 − 1)(115.2)
b 1 − β1 1.959 − 0
t= = = 6.949 (Excel: t = 6.95, p–value = 0.) There is evidence of a linear
s b1 .2819
relationship between food budget and household income.
526
⎛ s 2xy ⎞ ⎛ (−10.78) 2 ⎞
16.38 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (30 − 1)⎜11.24 − ⎟ = 230.9
⎜ s x ⎟⎠ ⎜ 35.47 ⎟
⎝ ⎝ ⎠
SSE 230.9
sε = = = 2.872 (Excel: s ε = 2.873 ).
n−2 30 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025, 28 = 2.048 or t < − t α / 2,n − 2 = − t .025, 28 = −2.048
sε 2.872
s b1 = = = .08955
(n − 1)s 2x (30 − 1)(35.47)
b 1 − β1 −.3039 − 0
t= = = −3.39 (Excel: t = –3.39, p–value = .0021.) There is sufficient evidence
s b1 .08955
to conclude that office rents and vacancy rates are linearly related.
⎛ s 2xy ⎞ ⎛ (6.020) 2 ⎞
16.39 SSE = (n − 1)⎜ s 2y − 2 ⎟ = (250 − 1)⎜ 71.95 − ⎟ = 17,010
⎜ s x ⎟⎠ ⎜ 9.966 ⎟
⎝ ⎝ ⎠
SSE 17,010
sε = = = 8.28 (Excel: s ε = 8.28).
n−2 250 − 2
H 0 : β1 = 0
H1 : β1 > 0
Rejection region: t > t α,n − 2 = t .05, 248 ≈ 1.645
sε 8.28
s b1 = = = .166
(n − 1)s 2x (250 − 1)(9.966
b 1 − β1 .604 − 0
t= = = 3.64 (Excel: t = 3.63, p–value = .00034/2 = .00017) There is enough
s b1 .166
evidence to conclude that height and income are positively linearly related.
s 2xy (.8258) 2
16.40 a R 2 = = = .0331 (Excel: R 2 = .0331) 3.31% of the variation in
s 2x s 2y (16.07)(1.283)
percentage of defectives is explained by the variation in aptitude test scores.
⎛ s 2xy ⎞ ⎛ (.8258) 2 ⎞
b. SSE = (n − 1)⎜ s 2y − 2 ⎟ = (45 − 1)⎜1.283 − ⎟ = 54.58
⎜ s x ⎟⎠ ⎜ 16.07 ⎟
⎝ ⎝ ⎠
SSE 54.58
sε = = = 1.127 (Excel: s ε = 1.127).
n−2 45 − 2
527
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025, 43 ≈ 2.014 or t < − t α / 2,n − 2 = − t .025, 43 ≈ −2.014
sε 1.127
s b1 = = = .04238
(n − 1)s 2x (45 − 1)(16.07)
b 1 − β1 .0516 − 0
t= = = 1.22 (Excel: t = 1.21, p–value = .2319) There is not enough evidence to
s b1 .04238
conclude that aptitude test scores and percentage of defectives are linearly related.
16.41 H0 : ρ = 0
H1 : ρ < 0
Rejection region: t < − t α 2,n − 2 = − t .05,58 ≈ −1.671
s xy − 81.78
r= = = −.1767
sxsy (59,153)(3.623)
n−2 60 − 2
t=r 2
= (−.1767) = −1.367 (Excel: t = –1.367, p–value = .0885) There is not
1− r 1 − (−.1767) 2
enough evidence to infer a negative linear relationship.
16.42 H0 : ρ = 0
H1 : ρ ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025,58 ≈ 2.000 or t < − t α / 2,n − 2 = − t .025,58 = −2.000
s xy 51.86
r= = = .5378
sxsy (193.9)(47.96)
n−2 60 − 2
t=r 2
= (.5378) = 4.86 (Excel: t = 4.86, p–value = 0) This result is identical
1− r 1 − (.5378) 2
to the one produced in Exercise 16.6.
16.43 H0 : ρ = 0
H1 : ρ ≠ 0
528
s xy 225.7
r= = = .4961
sxsy (115.2)(1,797)
n−2 150 − 2
t=r 2
= (.4961) = 6.95 (Excel: t = 6.95, p–value = 0.) There is evidence of a
1− r 1 − (.4961) 2
linear relationship between food budget and household income.
16.44 H0 : ρ = 0
H1 : ρ > 0
Rejection region: t > t α,n − 2 = t .05, 229 ≈ 1.645
s xy 20.55
r= = = .4438
sxsy (108.3)(19.80)
n−2 231 − 2
t=r 2
= (.4438) = 7.49 (Excel: t = 7.49, p–value = 0.) There is evidence of a
1− r 1 − (.4438) 2
positive linear relationship between cigarettes smoked and the number of sick days.
16.45
t = -1.45. There is not enough evidence to conclude that the more education one has the more one
watches news on the Internet. If anything the opposite appears to be the case.
529
16.46
t = 15.37, p-value = 5.43E-50 ≈ 0. There is enough evidence to infer that there is a positive linear
relationship between income and education.
16.47
t = 10.68, p-value = 3.94E-25 ≈ 0. There is enough evidence to infer that there is a positive linear
relationship between age and number of days watching national news on television.
530
16.48
t = 6.58, p-value = 6.06E-11 ≈ 0. There is enough evidence to infer that Age and intention to vote
are linearly related.
16.49
t = .69, p-value = .4915. There is not enough evidence to infer that there is a linear relationship
between income and time listening to news on the radio.
531
16.50
t = 7.80, p-value = 2.01E-14 ≈ 0. There is enough evidence to infer that there is a linear
relationship between income and position on the question should the government reduce income
differences between rich and poor.
16.51
532
b1 = 437, s b1 = 81. The 95% confidence interval estimate of β1: b1 ± t α / 2s b = 437 ± 1.96(81) =
1
437 ± 159
We estimate that the average increase in income for each additional year of age lies between $278
and $596.
16.52
t = -8.95, p-value = 1.21E-18 ≈ 0. There is enough evidence to conclude that more educated
people watch less television.
533
16.53
t = 17.48, p-value = 9.53E-65 ≈ 0. There is enough evidence to infer that more education leads to
higher income.
16.54 The prediction interval provides a prediction for a value of y. The confidence interval
estimator of the expected value of y is an estimator of the population mean for a given x.
16.55 ŷ = b 0 + b1 x g = 91.07 + .0582(90) = 14.35
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .05,10 = 1.812)
n (n − 1)s 2x
1 (90 − 51.08) 2
= 14.35 ± 1.812(1.347) 1 + + = 14.35 ± 2.747
12 (12 − 1)(749.7)
Lower prediction limit = 11.60, Upper prediction limit = 17.10 (Excel: 11.59, 17.09)
16.56 ŷ = b 0 + b1 x g = 475.2 − 39.17(8) = 161.8
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n − 2 = t .05,8 = 1.860)
n (n − 1)s 2x
1 (8 − 8.20) 2
= 161.8 ± 1.860(31.35) + = 161.8 ± 20.01
10 (10 − 1)(.24)
LCL = 141.8, UCL = 181.8 (Excel: 141.7, 182.0)
534
16.57 ŷ = b 0 + b1 x g = −24.72 + .9675(30) = 4.305
2
1 (x g − x)
a Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .05,13 = 1.771)
n (n − 1)s 2x
1 (30 − 31.47) 2
= 4.305 ± 1.771(3.825) 1 + + = 4.305 ± 7.007
15 (15 − 1)(47.98)
Lower prediction limit = –2.702, Upper prediction limit = 11.31 (Excel: –2.692, 11.32)
2
1 (x g − x)
b Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε +
n (n − 1)s 2x
1 (30 − 31.47) 2
= 4.305 ± 1.771(3.825) + = 4.305 ± 1.791
15 (15 − 1)(47.98)
LCL = 2.514, UCL = 6.096 (Excel: 2.524, 6.105)
16.58 ŷ = b 0 + b1 x g = −100,652 + 1,513(80) = 20,388
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .05,8 = 1.860)
n (n − 1)s 2x
1 (80 − 80.5) 2
= 20,388 ± 1.860(3,522) 1 + + = 20,388 ± 6,872
10 (10 − 1)(48.50)
Lower prediction limit = 13,516, Upper prediction limit = 27,260 (Excel: 13,518, 27,258)
16.59 ŷ = b 0 + b1 x g = 3.636 + .2675(36) = 13.27
2
1 (x g − x)
a Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025,58 ≈ 2.000)
n (n − 1)s 2x
1 (36 − 38) 2
= 13.27 ± 2.000(5.888) 1 + + = 13.27 ± 11.88
60 (60 − 1)(193.9)
Lower prediction limit =1.39, Upper prediction limit = 25.15 (Excel: 1.378, 25.15)
2
1 (x g − x)
b Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε +
n (n − 1)s 2x
1 (36 − 38) 2
= 13.27 ± 2.000(5.888) + = 13.27 ± 1.536
60 (60 − 1)(193.9)
LCL = 11.73, UCL = 14.81 (Excel: 11.73, 14.80)
535
16.60 ŷ = b 0 + b1 x g = 190.4 + 1.465(25) = 227.0
2
1 (x g − x)
a Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025, 48 ≈ 2.009)
n (n − 1)s 2x
1 (25 − 13.68) 2
= 227.0 ± 2.009(19.41) 1 + + = 227.0 ± 40.22
50 (50 − 1)(59.32)
b ŷ = b 0 + b1 x g = 190.4 + 1.465(12) = 208.0
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n − 2 = t .005, 48 = 2.678)
n (n − 1)s 2x
1 (12 − 13.68) 2
= 208.0 ± 2.678(19.41) + = 208.0 ± 7.527
50 (50 − 1)(59.32)
LCL = 200.5, UCL = 215.5 (Excel: 200.4, 215.5)
16.61 ŷ = b 0 + b1 x g = 7.46 + .0899(50) = 11.96
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n −2 = t .05,227 ≈ 1.645)
n (n − 1)s 2x
1 (50 − 45.49) 2
= 11.96 ± 1.645(6.47) + = 11.96 ± .77
229 (229 − 1)(107.51)
LCL = 11.19, UCL = 12.73 (Excel: 11.19, 12.73)
16.62 ŷ = b 0 + b1 x g = 30.64 – .1169(25) = 27.72
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025,78 ≈ 1.990)
n (n − 1)s 2x
1 (25 − 37.29) 2
= 27.72 ± 1.990(1.813) 1 + + = 27.72 ± 3.707
80 (80 − 1)(55.11)
16.63 ŷ = b 0 + b1 x g = 7.286 + .1898(30) = 12.98
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025, 229 ≈ 1.960)
n (n − 1)s 2x
536
1 (30 − 37.64) 2
= 12.98 ± 1.960(3.996) 1 + + = 12.98 ± 7.858
231 (231 − 1)(108.3)
16.64 ŷ = b 0 + b1 x g = 23.10 + 5.347(5) = 49.84
2
1 (x g − x)
a Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025,83 ≈ 1.990)
n (n − 1)s 2x
1 (5 − 4.885) 2
= 49.84 ± 1.990(11.10) 1 + + = 49.84 ± 22.22
85 (85 − 1)(4.270)
Lower prediction limit = 27.62, Upper prediction limit = 72.06 (Excel:27.62, 72.06)
b ŷ = b 0 + b1 x g = 23.10 + 5.347(2) = 33.79
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε +
n (n − 1)s 2x
1 (2 − 4.885) 2
= 33.79 ± 1.990(11.10) + = 33.79 ± 4.131
85 (85 − 1)(4.270)
LCL = 29.66, UCL = 37.92 (Excel: 29.67, 37.93)
16.65 ŷ = b 0 + b1 x g = 4,040 + 44.97(50) = 6,289
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n − 2 = t .025,38 ≈ 2.021)
n (n − 1)s 2x
1 (50 − 53.93) 2
= 6,289 ± 2.021(3,287) + = 6,289 ± 1,062
40 (40 − 1)(688.2)
LCL = 5,227, UCL = 7,351 (Excel: LCL = 5,224, UCL = 7,352)
16.66 ŷ = b 0 + b1 x g = 29.39 – .00138(500) = 28.70
2
1 (x g − x)
n (n − 1)s 2x
1 (500 − 1,199) 2
= 28.70 ± 2.660(1.889) 1 + + = 28.70 ± 5.404
60 (60 − 1)(59,153)
537
16.67 ŷ = b 0 + b1 x g = 458.4 + 64.05(5) = 778.65
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n − 2 = t .05,198 ≈ 1.653)
n (n − 1)s 2x
1 (5 − 4.75) 2
= 778.65 ± 1.653(191.1) + = 778.65 ± 22.48
200 (200 − 1)(4.84)
LCL = 756.17, UCL = 801.13 (Excel: 756.27,801.57)
16.68 ŷ = b 0 + b1 x g = 153.9 + 1.959(50) = 251.9
2
1 (x g − x)
n (n − 1)s 2x
1 (50 − 59.42) 2
= 251.9 ± 1.656(36.93) 1 + + = 251.9 ± 61.52
150 (150 − 1)(115.2)
16.69 ŷ = b 0 + b1 x g = 20.64 – .3039(10) = 17.60
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025, 28 = 2.048)
n (n − 1)s 2x
1 (10 − 11.33) 2
= 17.60 ± 2.048(2.872) 1 + + = 17.60 ± 5.984
30 (30 − 1)(35.47)
Lower prediction limit = 11.62 , Upper prediction limit = 23.58 (Excel: 11.61, 23.59)
16.70 a ŷ = b 0 + b1 x g = 17.94 + .604(72) = 61.43
2
1 (x g − x)
Confidence interval: ŷ ± t α / 2,n − 2 s ε + (where t α / 2,n − 2 = t .025, 248 ≈ 1.96)
n (n − 1)s 2x
1 (72 − 68.95) 2
= 61.43 ± 1.96(8.28 + = 61.43 ± 1.43
250 (250 − 1)(9.966)
Lower confidence limit = 60.00, Upper confidence limit = 62.86 (Excel: 59.99, 62.86)
b ŷ = b 0 + b1 x g = 17.94 + .604(66) = 57.80
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2,n − 2 = t .025, 248 ≈ 1.96)
n (n − 1)s 2x
538
1 (66 − 68.95) 2
= 57.80 ± 1.96(8.28 1 + + = 57.80 ± 16.29
250 (250 − 1)(9.966)
16.71 ŷ = b 0 + b1 x g = 89.81 + .0514(75) = 93.64
2
1 (x g − x)
Confidence interval estimate: ŷ ± t α / 2,n − 2 s ε + where t α / 2,n −2 = t .025,43 ≈ 2.014
n (n − 1)s 2x
1 (75 − 79.47) 2
= 93.64 ± 2.014(1.127) + = 93.64 ± .510
45 (45 − 1)(16.07)
LCL = 93.13, UCL = 94.15 (Excel: 93.15, 94.17)
16.72 16.71 ŷ = b 0 + b1 x g = 89.81 + .0514(80) = 93.92
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + where t α / 2,n −2 = t .05,43 ≈ 1.679
n ( n − 1)s 2x
1 (80 − 79.47) 2
= 93.92 ± 1.679(1.127) 1 + + = 93.92 ± 1.91
45 ( 45 − 1)(16.07)
16.73
Lower prediction limit = 0 (increased from -51.26), Upper prediction limit = 154.03
539
16.74
LCL = 16,466, UCL = 21,657
16.75
LCL = 3.44, UCL = 3.89
540
16.76
Lower prediction limit = 0 (increased from -83.09), Upper prediction limit = 204.81
16.77
Lower prediction limit =0 (increased from -32,759), Upper prediction limit = 117,916
541
16.78
LCL = 3.15, UCL = 3.40
16.79
Lower prediction limit = 0 (increased from -30,183), Upper prediction limit = 126,237
542
16.80
Lower prediction limit = 0 (increased from -.15), Upper prediction limit = 8.38
16.81 a x i yi x i2 yi2 x i yi
–5 15 25 225 –75
–2 9 4 81 –18
0 7 0 49 0
3 6 9 36 18
4 4 16 16 16
7 1 49 1 7
Total 7 42 103 408 –52
n n n n
∑ xi = 7 ∑ y i = 42 ∑ x i2 = 103 ∑x y i i = –52
i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑y
i =1
xi
i =1
i ⎥
⎥ 1 ⎡ (7)(42 ⎤
s xy = ∑
n − 1 ⎢⎢ i =1
x i yi −
n ⎥ = 6 − 1 ⎢− 52 − 6 ⎥ = −20.20
⎣ ⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ x i ⎥
⎟
⎢ n ⎜ ∑
⎟ ⎥
(7) 2 ⎤
⎢ x i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢103 −
1 ⎡
s 2x = ∑ ⎥ = 18.97
n − 1 ⎢ i =1 n ⎥ 6 − 1 ⎣⎢ 6 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s xy −20.2
b1 = = = −1.065
s 2x 18.97
x=
∑x i
=
7
= 1.167
n 6
543
y=
∑y i
=
42
= 7.000
n 6
b 0 = y − b1x = 7.000 – (–1.065)(1.167) = 8.253
ŷ = 8.253 – 1.065x
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ y i ⎥
⎟
⎢ n ⎜ ∑ ⎟ ⎥
(42) 2 ⎤
⎢ y i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢408 −
1 ⎡
b, c, &d s 2y = ∑ ⎥ = 22.80
n − 1 ⎢ i =1 n ⎥ 6 − 1 ⎣⎢ 6 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎛ s 2xy ⎞ ⎛ (20.20) 2 ⎞
SSE = (n − 1)⎜ s 2y − 2 ⎟ = (6 − 1)⎜ 22.80 − ⎟ = 6.451
⎜ s x ⎟⎠ ⎜ 18.97 ⎟
⎝ ⎝ ⎠
SSE 6.451
sε = = = 1.270 (Excel: s ε = 1.268)
n−2 6−2
xi yi ŷ i = 8.253 − 1.065x e i = y i − ŷ i ei / sε
–5 15 13.58 1.42 1.118
–2 9 10.38 –1.38 –1.087
0 7 8.253 –1.253 –.987
3 6 5.058 .942 .742
4 4 3.993 .007 .0055
7 1 .798 .202 .159

There are no outliers.
16.82 xi yi ŷ i = 9.107 − .0582x e i = y i − ŷ i

23 9.6 10.45 –.85
46 11.3 11.78 –.48
60 12.8 12.60 .20
54 9.8 12.25 –2.45
28 8.9 10.74 –1.84
33 12.5 11.03 1.47
25 12.0 10.56 1.44
31 11.4 10.91 .49
36 12.6 11.20 1.40
88 13.7 14.23 –.53
90 14.4 14.35 .06
99 15.9 14.87 1.03
16.83 xi yi ŷ = 475.2 – 39.17x e i = y i − ŷ i

8.5 115 142.3 –27.3
7.8 111 169.7 –58.7
7.6 185 177.5 7.5
7.5 201 181.4 19.6
8.0 206 161.8 44.2
8.4 167 146.2 20.8
544
8.8 155 130.5 24.5
8.9 117 126.6 –9.6
8.5 133 142.3 –9.3
8.0 150 161.8 –11.8
16.84 a & b xi yi ŷ = – 24.72 + .9675x e i = y i − ŷ i

42 18 15.92 2.09
34 6 8.18 –2.18
25 0 –.53 .53
35 –1 9.14 –10.14
37 13 11.08 1.92
38 14 12.05 1.96
31 7 5.27 1.73
33 7 7.21 –.21
19 –9 –6.34 –2.66
29 8 3.34 4.66
38 8 12.05 –4.05
28 5 2.37 2.63
29 3 3.34 –.34
36 14 10.11 3.89
18 –7 –7.31 .31
Plot of Residuals vs Predicted
6
4
2
0
Residuals
-10 -5 -2 0 5 10 15 20
-4
-6
-8
-10
-12
Predicted
16.85 xi yi ŷ = –100,652 + 1,513x e i = y i − ŷ i

80 20,533 20,388 145
68 1,439 2,232 –793
78 13,829 17,362 –3,533
79 21,286 18,875 2,411
87 30,985 30,979 6
74 17,187 11,310 5,877
86 30,240 29,466 774
92 37,596 38,544 –948
77 9,610 15,849 –6,239
84 28,742 26,440 2,302
545
The histograms drawn below are of the standardized residuals, which make it easier to see
whether the shape is extremely nonnormal. It also makes it easier to identify outliers. The shape of
the resulting histogram is identical to the histogram of the residuals using the equivalent class
limits.
16.86 b & c
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standardized Residuals
Because the histogram is approximately bell shaped the errors appear to be normally distributed.
There are two residuals whose absolute value exceeds 2.0.
20
15
10
5
Residuals
0
-5 0 5 10 15 20 25
-10
-15
-20
Predicted
There is no indication of heteroscedasticity.
546
16.87 a
Histogram
Frequency
40
20
0
-3 -2 0 1 1 2 3
Standardized Residuals
The histogram is not bell shaped. However, residuals do not appear to be extremely nonnormal.
c
60
50
40
30
20
Residuals
10
0
-10180 190 200 210 220 230 240
-20
-30
-40
-50
Predicted
There is no clear indication of heteroscedasticity.
16.88
The error variable appears to be normally distributed.
547
The variance of the error variable is constant.
16.89 b & c
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
The error variable appears to be normally distributed. There are three observations whose
standardized residuals are greater than 2.0.
548
d
2
Residuals
0
24 25 26 27 28 29
-2
-4
-6
Predicted
16.90
Histogram
Frequency
100
50
0
-3 -2 -1 0 1 2 More
Standard Residuals
15
10
5
Residuals
0
8 12 16 20
-5
-10
-15
Predicted
549
16.91b
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
The errors appear to be normally distributed.

c There are no outliers.
d
40
30
20
Residuals
10
0
20 30 40 50 60 70 80
-10
-20
-30
Predicted
The variance of the error variable appears to decrease somewhat as the predicted values increase.
However, the effect is not large enough to be a problem.
16.92
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
550
8000
6000
4000
2000
Residuals
0
-20004000 5000 6000 7000 8000 9000
-4000
-6000
-8000
Predicted
There is no clear sign of heteroscedasticity.
16.93
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
4
Residuals
0
27 27.5 28 28.5
-2
-4
-6
Predicted
551
16.94
Histogram
Frequency 100
0
-‐3 -‐2 -‐1 0 1 2 3
Standard Residuals
The error variable appears to be normal.

600
400
200
0
400 500 600 700 800 900 1000
-‐200
-‐400
-‐600
-‐800
16.95
Histogram
Frequency
100
50
0
-3 -2 -1 0 1 2 3
Standard Residuals
552
150
100
50
Residuals
0
200 250 300 350
-50
-100
-150
Predicted
16.96
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
2
Residuals
0
14 18 22
-2
-4
-6
Predicted
553
16.97
Histogram
Frequency 100
0
-‐3 -‐2 -‐1 0 1 2 3
Standard Residuals
25
20
Residuals vs Predicted
15
10
5
0
-‐5 54 56 58 60 62 64 66
-‐10
-‐15
-‐20
-‐25
16.98
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
554
3
2
1
Residuals
0
93.2 93.4 93.6 93.8 94 94.2 94.4
-1
-2
-3
-4
Predicted
s xy 74.02
16.99 a b1 = = = 21.33, b 0 = y − b1x = 384.81 – 21.33(4.12) = 296.93
s 2x 3.47
b On average each additional ad generates 21.33 customers.
⎛ s 2xy ⎞ ⎛ (74.02) 2 ⎞
c SSE = (n − 1)⎜ s 2y − 2 ⎟ = (26 − 1)⎜18,552 − ⎟ = 424,326
⎜ s x ⎟⎠ ⎜ 3.47 ⎟
⎝ ⎝ ⎠
SSE 424,326
sε = = = 132.97 (Excel: s ε = 132.96).
n−2 26 − 2
H 0 : β1 = 0
H1 : β1 > 0
Rejection region: t > t α,n − 2 = t .05, 24 = 1.711
sε 132.97
s b1 = = = 14.28
(n − 1)s 2x (26 − 1)(3.47)
b 1 − β1 21.33 − 0
t= = = 1.49 (Excel: t = 1.50, p–value = .1479/2 = .0740.) There is not enough
s b1 14.28
evidence to conclude that the larger the number of ads the larger the number of customers.
s 2xy (74.02) 2
d R2 = = = .0851(Excel: R 2 = .0852). There is a weak linear relationship
s 2x s 2y (3.47)(18,552)
between the number of ads and the number of customers.

e The linear relationship is too weak for the model to produce predictions.
555
s xy 936.82
16.100 a b1 = = = 2.47 b 0 = y − b1x = 395.21 – 2.47(113.35) = 115.24.
s 2x 378.77
b b1 = 2.47; for each additional month of age, repair costs increase on average by $2.47.
b 0 = 114.85 is the y-intercept.
s 2xy (936.82) 2
c R2 = = = .5659 (Excel: R 2 = .5659) 56.59% of the variation in repair
s 2x s 2y (378.77)(4,094.79)
costs s explained by the variation in ages.
⎛ s 2xy ⎞ ⎛ (936.82) 2 ⎞⎟

d SSE = (n − 1)⎜ s 2y − 2 ⎟ = (20 − 1)⎜ 4,094.79 − = 33,777
⎜ s ⎟ ⎜ 378 .77 ⎟
⎝ x ⎠ ⎝ ⎠
SSE 33,777
sε = = = 43.32 (Excel: s ε = 43.32).
n−2 20 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t .025,18 = 2.101or t < − t α / 2, n − 2 = − t .025,18 = −2.101
sε 43.32
s b1 = = = .511
(n − 1)s 2x (20 − 1)(378.77)
b 1 − β1 2.47 − 0
t= = = 4.84 (Excel: t = 4.84, p–value = .0001. There is enough evidence to infer
s b1 .511
that repair costs and age are linearly related.

e ŷ = b0 + b1xg = 115.24 + 2.47(120) = 411.64
2
1 (x g − x)
Prediction interval: ŷ ± t α / 2,n −2 s ε 1 + + (where t α / 2, n − 2 = t.025,18 = 2.101)
n (n − 1)s 2x
1 (120 − 113.35) 2
= 411.64 ± 2.101(43.32) 1 + + = 411.64 ± 93.54
20 (20 − 1)(378.77)
Lower prediction limit = 318.1, upper prediction limit = 505.2 (Excel: 318.1, 505.2)
s xy 2538
16.101 a b1 = = = .123 b 0 = y − b1x = 318.60 –.123(300) = 281.7.
s 2x 20,690
The slope is .123, which tells us that for each additional unit of fertilizer, corn yield increases on
average by .123. The y-intercept is 281.7, which has no real meaning.
556
⎛ s 2xy ⎞ ⎛ (2,538)2 ⎞⎟
b SSE = ( n − 1)⎜ s 2y − 2 ⎟ = (30 − 1)⎜ 5,230 − = 142,641
⎜ s x ⎟⎠ ⎜ 20,690 ⎟⎠
⎝ ⎝
SSE 142,641
sε = = = 71.37 (Excel: s ε = 71.38).
n−2 30 − 2
H 0 : β1 = 0
H1 : β1 ≠ 0
Rejection region: t > t α / 2, n − 2 = t.025, 28 = 2.048 or t < − t α / 2, n − 2 = −t.025,28 = −2.048
sε 71.37
s b1 = = = .0921
(n − 1)s 2x (30 − 1)(20,690)
b 1 − β1 .123 − 0
t= = = 1.34 (Excel: t = 1.33, p–value = .1938. There is not enough evidence to
s b1 .0921
infer a linear relationship between amount of fertilizer and corn yield.
s 2xy (2,538) 2
c R2 = = = .0595 ( Excel: R 2 = .0595) 5.95% of the variation in corn yield
s 2x s 2y (20,690)(5,230)
is explained by the variation in amount of fertilizer.

d The model is too poor to be used to predict.
16.102a H0 : ρ = 0
H1 : ρ ≠ 0
A B
1 Correlation
2
3 Tar and Nicotine
4 Pearson Coefficient of Correlation 0.9766
5 t Stat 21.78
6 df 23
7 P(T<=t) one tail 0
8 t Critical one tail 1.7139
9 P(T<=t) two tail 0
10 t Critical two tail 2.0687
r = .9766, t = 21.78, p–value = 0. There is sufficient evidence to infer that levels of tar and nicotine
are linearly related.
b H0 : ρ = 0
H1 : ρ ≠ 0
557
A B
1 Correlation
2
3 Nicotine and CO
5 t Stat 11.76
6 df 23
r = .9259, t = 11.76, p–value = 0. There is sufficient evidence to infer that levels of nicotine and
carbon monoxide are linearly related.
16.103 H0 : ρ = 0
H1 : ρ > 0
Rejection region: t > t α, n − 2 = t .05, 428 = 1.645 or
s xy 255,877
r= = = .5540 (Excel: .5540)
s xs y (99.11)(2,152,602,614)
n−2 430 − 2
t=r 2
= (.5540) = 13.77 (Excel: t = 13.77, p–value = 0). There is enough
1− r 1 − (.5540) 2
evidence of a positive linear relationship. The theory appears to be valid.
16.104 H0 : ρ = 0
H1 : ρ ≠ 0
Rejection region: t > t α / 2,n − 2 = t .025, 48 ≈ 2.009 or t < − t α / 2,n − 2 = − t .025, 48 = −2.009
s xy 13.08
r= = = .3985 (Excel: .3984)
sxsy (90.97)(11.84)
n−2 50 − 2
t=r 2
= (.3985) = 3.01 (Excel: t = 3.01, p–value = .0042). There is enough
1− r 1 − (.3985) 2
evidence of a linear relationship. The theory appears to be valid.
16.105 H0 : ρ = 0
H1 : ρ > 0
558
A B C D
1 Correlation
2
3 Fund and Gold
5 t Stat 6.63
6 df 26
t = 6.63, p-value = 0. There is enough evidence to conclude that there is a positive linear
relationship between the value of the fund and the price of gold.
16.106 H0 : ρ = 0
H1 : ρ > 0
A B C D
1 Correlation
2
3 Time and Sales
5 t Stat 1.67
6 df 33
7 P(T<=t) one tail 0.0522
9 P(T<=t) two tail 0.1044
t = 1.67, p-value = .0522. There is not enough evidence to infer that when the times between
movies increase so do sales.
559
16.107 a
b. H0 : ρ = 0
H1 : ρ ≠ 0
t = 7.60, p-value = 0. There is enough evidence to infer that index finger length and height are
linearly related.
560
c.
Lower prediction limit = 150.5, Upper prediction limit = 182.5
16.108 H0 : ρ = 0
H1 : ρ ≠ 0
A B
1 Correlation
2
3 Hours and GPA
4 Pearson Coefficient of Correlation -0.5748
5 t Stat -9.88
6 df 198
r = t = –9.88, p–value = 0. There is enough evidence of a linear relationship between time spent at
part time jobs and grade point average.
16.109 H0 : ρ = 0
H1 : ρ ≠ 0
561
A B
1 Correlation
2
3 Times and Amount
5 t Stat 29.51
6 df 498
t = 29.51, p–value = 0. There is overwhelming evidence of a linear relationship between listening
times and amounts spent on music.
Case 16.1 a
A B C D E F
1 SUMMARY OUTPUT
2
3 Regression Statistics
4 Multiple R 0.9896
5 R Square 0.9794
6 Adjusted R Square 0.9787
7 Standard Error 355.6
8 Observations 32
9
10 ANOVA
11 df SS MS F Significance F
12 Regression 1 180,066,717 180,066,717 1424 0.0000
13 Residual 30 3,793,289 126,443
14 Total 31 183,860,006
15
16 Coefficients Standard Error t Stat P-value
17 Intercept 16.23 114.70 0.14 0.8884
18 A-Park 0.693 0.0184 37.74 0.0000
The regression equation is ŷ = 16.23 + .693x. This equation was used to predict museum
attendance when it was closed (observations 33 to 179). The sum of the predictions is 785,009.
562
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9909
5 R Square 0.9819
8 Observations 26
9
10 ANOVA
12 Regression 1 426,295,375 426,295,375 1299 0.0000
13 Residual 24 7,875,875 328,161
14 Total 25 434,171,250
15
17 Intercept 459.5 295.4 1.56 0.1330
18 A-Park 0.970 0.0269 36.04 0.0000
The regression equation is ŷ = 459.5 + .970x. This equation was used to predict museum
attendance when it was closed (observations 33 to 179). The sum of the predictions is 1,162,994.
c The predicted lost revenue should be based on the regression using the first 32 weeks. Multiply
785,009 by the price of tickets and subtract fixed costs to produce the amount the insurance
company should pay the museum.
Case 16.2
Regression using the best 6 OACs:
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.4883
5 R Square 0.2385
8 Observations 363
9
10 ANOVA
12 Regression 1 77.78 77.78 113.04 0.0000
13 Residual 361 248.39 0.69
14 Total 362 326.17
15
17 Intercept -5.35 1.31 -4.08 0.0001
18 Best-6 0.15 0.01 10.63 0.0000
563
t = 10.63, p–value = 0. There is evidence of a linear relationship between the average of the best 6
OACs and university GPA. R 2 = .2385, sε = .8295
Regression using the best 4 OACs plus English and calculus:

A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.5924
5 R Square 0.3509
8 Observations 363
9
10 ANOVA
12 Regression 1 114.46 114.46 195.17 0.0000
13 Residual 361 211.71 0.59
14 Total 362 326.17
15
17 Intercept -3.32 0.85 -3.89 0.0001
18 B4+E+C 0.14 0.01 13.97 0.0000
t = 13.97, p–value = 0; there is evidence of a linear relationship between the average of the best 4
OACs plus English and calculus and university GPA. R 2 = .3509, sε = .7658.
The second model fits better (higher coefficient of determination and lower standard error of
estimate) and as such is likely to be a better predictor of university GPA.
564
Appendix 16
A16.1 t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
A B
1 Correlation
2
3 Weight and B/A Level
5 t Stat 3.19
6 df 48
r = .4177, t = 3.19, p–value = .0026. There is sufficient evidence to infer that weight and blood–
alcohol level are related.

H 0 : µ1 = µ 2 = µ 3 = µ 4
A B C D E F G
40 ANOVA
42 Rows 3708.8 28 132.46 8.63 6.52E-15 1.61
43 Columns 997.0 3 332.33 21.64 1.77E-10 2.71
44 Error 1289.8 84 15.35
45
46 Total 5995.5 115
F = 21.64; p-value = 0. There is enough evidence to conclude that there are differences in the
decrease in test scores between the four types of breakfast meals.
A16.3a Chi-squared goodness-of-fit test (the percentages must be converted to actual and expected
values and we must include those who did not have cancer)
H 0 : p1 = 143 / 420,000, p 2 = 9 / 420,000, p 3 = 80 / 420,000, p 4 = 52 / 420,000,

p 5 = 57 / 420,000, p 6 = 12 / 420,000, p 7 = 13 / 420,000, p 8 = 419.634 / 420,000
565

A B C D
1 Actual Expected
2 135 143
3 7 9
4 77 80
5 32 52
6 42 57
7 8 12
8 13 13
9 419686 419634 p-value = 0.0515
p-value = .0515. There is not enough evidence to conclude that there is a relationship between cell
phone use and cancer.
b The data are observational. Even if we regard the statistical result as significant we cannot
automatically infer that cell phone use causes cancer. Additionally, an examination of the actual
and expected values reveals that in all 7 types of cancers the actual values are less than or equal to
the expected values, indicating that (if anything) cell phone use prevents cancer.
A16.4 t-test of ρ or t-test of β1

H0 : ρ = 0
H1 : ρ > 0

A B C D
1 Correlation
2
3 Age and Duration
5 t Stat 7.90
6 df 138
t = 7.90; p-value = 0. There is overwhelming evidence to infer that the older the patient the longer
it takes for the symptoms to disappear?
A16.5a Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
566

A B C
2
3 Home Outside
4 Mean 59.21 54.91
5 Variance 102.03 88.28
9 df 346
10 t Stat 4.06
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0001
t = 4.06, p-value = 0. There is enough evidence to infer that men whose wives stay at home earn
more than men whose wives work outside the home.
b It may be that men whose wives stay at home work harder, and thus earn more.
A16.6 Question 1: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 US Days Canada Days
4 Mean 26.98 29.44
5 Variance 55.90 56.82
9 df 598
10 t Stat -4.00
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0001
t = –4.00, p-value = 0. There is enough evidence to indicate that recovery is faster in the United
States.
Question 2: z-tests of p1 − p 2 (case 1)
567

H0: (p1 – p2) = 0
H1: (p1 – p2) < 0
A B C D
2
3 U.S. Canada
7 z Stat -1.55
10 P(Z<=z) two-tail 0.1218
z = –1.55, p-value = .0609. There is not enough evidence to infer that recovery is faster in the
United States.
6 months after heart attack:

A B C D
2
3 U.S. Canada
7 z Stat 0.43
10 P(Z<=z) two-tail 0.6708
z = .43, p-value = 1 - .3354 = .6646. There is no evidence to infer that recovery is faster in the
United States.
568

12 months after heart attack
A B C D
2
3 U.S. Canada
7 z Stat 0.26
10 P(Z<=z) two-tail 0.7968
z = .26, p-value = 1 – .3984 = .6016. There is no evidence to infer that recovery is faster in the
United States.

H 0 : The two variables are independent
A B C D E F
1 Contingency Table
2
3 Favored
4 Result 1 2 3 TOTAL
5 1 31 25 17 73
6 2 46 16 19 81
7 3 27 7 15 49
8 4 16 2 3 21
9 TOTAL 120 50 54 224
10
11
13 df 6
14 p-value 0.0365
χ 2 = 13.4477, p-value = .0365. There is enough evidence to infer that Pro-Line's forecasts are
related to outcomes and thus, can be useful to bettors.
H0 : µD = 0
H1 : µ D < 0
569

A B C
2
3 No-Slide Slide
4 Mean 3.73 3.78
5 Variance 0.0653 0.0727
9 df 24
10 t Stat -3.04
11 P(T<=t) one-tail 0.0028
13 P(T<=t) two-tail 0.0057
t = –3.04, p-value = .0028. There is overwhelming evidence to indicate that sliding is slower.
A16.9 z-test of p1 − p 2 (case 1) (The data were unstacked prior to applying the z-test.)
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Optimist Pessimist
7 z Stat 4.26
8 P(Z<=z) one tail 0
10 P(Z<=z) two-tail 0
z = 4.26, p-value = 0. There is sufficient evidence that pessimists are less likely to survive than
optimists.
A16.10 Simple linear regression with cholesterol reduction (Before – After) as the dependent
variable
a t-test of β1 or test of ρ
H 0 : β1 = 0
H1 : β1 ≠ 0
570

We used the t-test of β1 because parts (b) and (c) use the regression equation to predict and
estimate.
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7138
5 R Square 0.5095
8 Observations 50
9
10 ANOVA
12 Regression 1 5528 5528.5 49.87 5.92E-09
13 Residual 48 5322 110.9
14 Total 49 10850
15
17 Intercept 2.05 3.94 0.52 0.6051
18 Exercise 0.0909 0.0129 7.06 5.92E-09
t = 7.06; p-value = 0. There is overwhelming evidence to infer that exercise and cholesterol
reduction are related.
b. Prediction interval
A B C
1 Prediction Interval
2
3 Reduction
4
5 Predicted value 11.14
6
8 Lower limit -10.76
9 Upper limit 33.05
10
11 Interval Estimate of Expected Value
12 Lower limit 5.54
13 Upper limit 16.75
The cholesterol reduction is predicted to fall between −10.76 and 33.05.
571

c Confidence interval estimator of the expected value of y
A B C
2
3 Reduction
4
6
8 Lower limit -8.83
9 Upper limit 34.76
10
12 Lower limit 7.79
We estimate that the mean reduction in cholesterol lies between 7.79 and 18.14.
A16.11 Instructors: Unequal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
t = −2.40; p-value = .0087. There is enough evidence to conclude that the salaries of instructors at
publicly-funded colleges and universities are less than the salaries of instructors at private colleges
and universities.
572

Assistant professors: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Public-Assistant Private-Assistant
4 Mean 54.08 59.37
5 Variance 26.10 32.84
9 df 265
10 t Stat -7.98
11 P(T<=t) one-tail 2.18E-14
13 P(T<=t) two-tail 4.36E-14
t = −7.98; p-value = 0. There is overwhelming evidence to conclude that the salaries of assistant
professors at publicly-funded colleges and universities are less than the salaries of assistant
professor at private colleges and universities.
Associate professors: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
Two-tail F test: F = 1.00, p-value = 1.00; use equal-variances t-test
A B C
2
3 Public-Associate Private-Associate
4 Mean 64.45 71.07
5 Variance 30.96 30.96
9 df 320
10 t Stat -10.69
11 P(T<=t) one-tail 2.65E-23
13 P(T<=t) two-tail 5.31E-23
573

t = −10.69; p-value = 0. There is overwhelming evidence to conclude that the salaries of associate
professors at publicly-funded colleges and universities are less than the salaries of associate
professor at private colleges and universities.
Professors: Unequal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Public-Professor Private-Professor
4 Mean 88.89 107.39
5 Variance 49.14 74.99
8 df 310
9 t Stat -23.51
10 P(T<=t) one-tail 3.61E-71
12 P(T<=t) two-tail 7.22E-71
t = −23.51; p-value = 0. There is overwhelming evidence to conclude that the salaries of
professors at publicly-funded colleges and universities are less than the salaries of professors at
private colleges and universities.
A16.12a One-way analysis of variance

H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 1.65 2 0.823 0.1332 0.8753 3.0259
13 Within Groups 1847.2 299 6.18
14
15 Total 1848.9 301
F = .1332; p-value = .8753. There is no evidence to infer that there are differences between the
three groups of patients.
574

b One-way analysis of variance
H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 247.0 2 123.48 15.81 2.96E-07 3.03
13 Within Groups 2334.6 299 7.81
14
15 Total 2581.6 301
F = 15.81; p-value = 0. There is overwhelming evidence to conclude that there are differences
between the three groups of patients.
Multiple comparisons
A B C D E
2
3 LSD Omega
5 Group 1 After Group 2 After 0.099 0.949 0.922
6 Group 3 After -1.867 0.942 0.922
7 Group 2 After Group 3 After -1.965 0.954 0.922
Group 3 differs from both group 1 and group 2. Groups 1 and 2 do not differ.
c. The test assures researchers that the three groups of patients were very similar prior to
treatments.
A16.13 a. One-way analysis of variance

H 0 : µ1 = µ 2 = µ 3 = µ 4 = µ 5 = µ 6
A B C D E F G
13 ANOVA
15 Between Groups 527,465 5 105,493 4.43 0.0015 2.3538
16 Within Groups 1,571,667 66 23,813
17
18 Total 2,099,132 71
F = 4.43; p-value = .0015. There is enough evidence to infer that differences exist between the six
groups.
575

b. Two-factor analysis of variance
A B C D E F G
29 ANOVA
31 Sample 303,247 2 151,623 6.37 0.0030 3.1359
32 Columns 190,139 1 190,139 7.98 0.0062 3.9863
33 Interaction 34,080 2 17,040 0.72 0.4927 3.1359
34 Within 1,571,667 66 23,813
35
36 Total 2,099,132 71
Test for interaction: F = .72; p-value = .4927. There is no evidence of interaction.
Test for gender (columns): F = 7.98; p-value = .0062. There is enough evidence to conclude that
there are differences in cash offers between males and females.
Test for age: F = 6.37; p-value = .0030. There is enough evidence to conclude that there are
differences in cash offers between the three age groups.
A16.14 a Chi-squared test of a contingency table

H 0 : The two variables (year and party) are independent
A B C D E F
1 Contingency Table
2
3 1990 1996 2000 2004 TOTAL
4 Democrats 154 161 159 152 626
5 Republicans 99 100 97 87 383
6 Other 22 42 56 60 180
7 TOTAL 275 303 312 299 1189
8
10 df 6
11 p-value 0.0037
χ 2 = 19.27; p-value = .0037. There is overwhelming evidence to infer that party affiliation in
Broward County changed over the four years.
b Chi-squared test of a contingency table

H 0 : The two variables (year and party) are independent
576

A B C D E F
1 Contingency Table
2
3 1990 1996 2000 2004 TOTAL
4 Democrats 173 157 136 146 612
5 Republicans 117 128 117 122 484
6 Other 25 43 56 63 187
7 TOTAL 315 328 309 331 1283
8
10 df 6
11 p-value 0.0009
χ 2 = 22.65; p-value = .0009. There is overwhelming evidence to infer that party affiliation in
Miami-Dade changed over the four years.
c Chi-squared test of a contingency table

H 0 : The two variables (County and party in 2004) are independent
A B C D
1 Contingency Table
2
3 Broward Miami-Dade TOTAL
4 Democrats 152 146 298
5 Republicans 87 122 209
6 Other 60 63 123
7 TOTAL 299 331 630
8
10 df 2
11 p-value 0.1085
χ 2 = 4.44; p-value = 1095. There is not enough evidence to infer that party affiliation differ
between Broward County and Miami-Dade County.
A16.15 t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
577

A B
1 Correlation
2
3 CO and NO3
5 t Stat 13.62
6 df 48
r = .8913, t = 13.62, p–value = 0; there is enough evidence to infer that the belief is correct.
A B C D
1 t-Estimate: Mean
2
3 Commute
4 Mean 24.54
6 LCL 23.64
7 UCL 25.45
Total time spent commuting by all workers:
LCL = 129,142,000 (23.64) = 3,052,916,880 minutes
UCL = 129,142,000 (25.45) = 3,286,663,900 minutes
A16.17 H0: p = .5
H1: p < .5
1= Success
578

z = 21.01. There is no evidence to conclude that less than 50% of Americans support gun laws. In
fact there is evidence to the contrary.
A16.18 H0: (µ1 - µ2) = 0

H1: (µ1 - µ2) ≠ 0
Two-tail F test: F = .814, p-value = .0383; Use unequal-variances t-test.
t = -12.98, p-value = 1.52E-34 ≈ 0. There is sufficient evidence to infer that Democrats and
Republicans differ In their positions on whether the government should reduce income differences
between rich and poor.
A16.19 H0: ρ = 0
H 1: ρ ≠ 0
t = 4.27, p-value = 0. There is enough evidence to conclude that there is a linear relationship
between income and the responses to the question about whether the government should improve
the living conditions of poor people
579

A16.20 H0: µ1 = µ2 = µ3 = µ4
H1: At least two means differ
F = .162, p-value = .9217. There is not enough evidence to conclude that there are differences
between union households with respect to their position on whether the government should
improve the living conditions of poor people.
A16.21 H0: (p1 – p2) = 0

H1: (p1 – p2) > 0
1 = Success
z = .865, p-value = .1934. There is not enough evidence to conclude that people who have taken
college-level science courses are more likely to answer the question correctly.
580

A16.22 H0: ρ = 0
H 1: ρ > 0
t = 1.87, p-value = .0308. There is enough evidence to infer that there is a positive linear
relationship between income and the number of people working for the company.
A16.23 Ch-squared test of a contingency table

H 0 : The two variables (PARTYID3 and FINAN1) are independent
χ 2 = 4.55, p-value = .1030. There is not enough evidence to conclude that differences exist
between people who went bankrupt and those who did not in their political affiliations.
A16.24 H0: ρ = 0
H 1: ρ > 0
581

t = 9.35, p-value = 0. There is sufficient evidence to infer that the more one works the greater the
income.
A16.25 H0: (p1 – p2) = 0

H1: (p1 – p2) < 0
1 = Success
z = -1.09, p-value = .1380. There is not enough vidence to infer that victims of robbery are less
likely to support gun laws.
A16.26 H0: ρ = 0
H 1: ρ > 0
582

relationship between education and occupation prestige score.
A16.27 H0: ρ = 0
H 1: ρ > 0
t = 1.77, p-value .0415. There is enough evidence to conclude that people with more education
spend more time reading newspapers.
A16.28 H0: ρ = 0
H 1: ρ ≠ 0
583

t = 16.45, p-value 0. There is enough evidence to conclude that the number of days watching
national news is linearly related to the number of days watching local news in the late afternoon or
early evening.
Case A16.2a t-tests of µ D
H0 : µD = 0
H1 : µ D > 0
Weight
A B C
2
3 Weight 1 Weight 2
4 Mean 79.89 78.62
5 Variance 255.54 251.09
9 df 32
10 t Stat 2.90
11 P(T<=t) one-tail 0.0034
13 P(T<=t) two-tail 0.0067
t = 2.90, p-value = .0034. There is enough evidence to infer that the program is a success in terms
of weight level.
584

Cholesterol
A B C
2
3 Choles 1 Choles 2
4 Mean 6.87 6.27
5 Variance 0.583 0.618
9 df 32
10 t Stat 4.83
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
t = 4.83, p-value = 0. There is enough evidence to infer that the program is a success in terms of
cholesterol level.
Fat intake
A B C
2
3 TotFat 1 TotFat 2
4 Mean 66.56 46.72
5 Variance 967.59 533.91
9 df 32
10 t Stat 4.70
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
fat intake.
585

Cholesterol intake
A B C
2
3 DietC 1 DietC 2
4 Mean 242.42 177.12
5 Variance 30618 13032
9 df 32
10 t Stat 2.29
11 P(T<=t) one-tail 0.0144
13 P(T<=t) two-tail 0.0288
t = 2.29, p-value = .0144. There is enough evidence to infer that the program is a success in terms
of cholesterol intake.
Calories from fat

A B C
2
3 PDCF 1 PDCF 2
4 Mean 36.54 30.82
5 Variance 56.72 49.71
9 df 32
10 t Stat 6.29
11 P(T<=t) one-tail 0.0000
13 P(T<=t) two-tail 0.0000
daily calories from fat.
b Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
586

Weight reduction
A B C
2
3 Female Male
4 Mean 1.32 1.22
5 Variance 1.54 11.17
9 df 31
10 t Stat 0.11
11 P(T<=t) one-tail 0.4551
13 P(T<=t) two-tail 0.9102
t = .31, p-value = .9102. There is no evidence to infer that gender is a factor in weight reduction.
Cholesterol reduction
A B C
2
3 Female Male
4 Mean 0.48 0.72
9 df 31
10 t Stat -0.94
11 P(T<=t) one-tail 0.1776
13 P(T<=t) two-tail 0.3551
t = –.94, p-value = .3551. There is not enough evidence to infer that gender is a factor in
cholesterol reduction.
587

Fat intake reduction
A B C
2
3 Female Male
4 Mean 14.50 24.88
5 Variance 229.91 906.76
9 df 31
10 t Stat -1.24
11 P(T<=t) one-tail 0.1125
13 P(T<=t) two-tail 0.2251
t = –1.24, p-value = .2251. There is not enough evidence to infer that gender is a factor in fat
intake reduction.
Cholesterol intake reduction

A B C
2
3 Female Male
4 Mean 38.46 90.56
5 Variance 5683 46984
9 df 31
10 t Stat -0.91
11 P(T<=t) one-tail 0.1848
13 P(T<=t) two-tail 0.3697
t = –.91, p-value = .3697. There is not enough evidence to infer that gender is a factor in
cholesterol intake reduction.
588

Calories from fat reduction
A B C
2
3 Female Male
4 Mean 6.88 4.62
5 Variance 18.69 34.29
9 df 31
10 t Stat 1.25
11 P(T<=t) one-tail 0.1103
13 P(T<=t) two-tail 0.2207
t = 1.25, p-value = .2207. There is not enough evidence to infer that gender is a factor in calories
from fat reduction.
c t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
Age and weight reduction
A B C D
1 Correlation
2
3 Age and Weight
5 t Stat -0.55
6 df 31
t = –.55, p-value = .5874. There is not enough evidence that age is a factor in weight reduction.
589

Age and cholesterol reduction
A B C D
1 Correlation
2
3 Age and Choles
5 t Stat 2.40
6 df 31
t = 2.40, p-value = .0226. There is enough evidence to infer that age is a factor in cholesterol
reduction.
Age and Fat intake reduction

A B C D
1 Correlation
2
3 Age and TotFat
5 t Stat -0.84
6 df 31
t = –.84, p-value = .4074. There is not enough evidence that age is a factor in fat intake reduction.
Age and Cholesterol intake reduction

A B C D
1 Correlation
2
3 Age and Dietc
5 t Stat -0.71
6 df 31
t = –.71, p-value = .4854. There is not enough evidence that age is a factor in cholesterol intake
reduction.
590

Age and Calories from fat reduction
A B C D
1 Correlation
2
3 Age and PDCF
5 t Stat -2.17
6 df 31
t = –2.17, p-value = .0380. There is enough evidence to infer that age is a factor in calories from
fat reduction.
591

592

Chapter 17
17.1
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.4924
5 R Square 0.2425
8 Observations 60
9
10 ANOVA
12 Regression 3 29,030 9,677 5.97 0.0013
13 Residual 56 90,694 1,620
14 Total 59 119,724
15
17 Intercept 51.39 23.52 2.19 0.0331
18 Lot size 0.700 0.559 1.25 0.2156
19 Trees 0.679 0.229 2.96 0.0045
20 Distance -0.378 0.195 -1.94 0.0577
a ŷ = 51.39 + .700x1 + .679x 2 − .378x3
b The standard error of estimate is s ε = 40.24. It is an estimate of the standard deviation of the
error variable.
c The coefficient of determination is R 2 = .2425; 24.25% of the variation in prices is explained by

the model.
d The coefficient of determination adjusted for degrees of freedom is .2019. It differs from
R 2 because it includes an adjustment for the number of independent variables.

e H 0: β 1 = β 2 = β 3 = 0
H1: At least one β i is not equal to zero
F = 5.97, p-value = .0013. There is enough evidence to conclude that the model is valid.
f b1 = .700; for each addition thousand square feet the price on average increases by .700 thousand
dollars provided that the other variables remain constant.

b 2 = .679; for each addition tree the price on average increases by .679 thousand dollars provided
that the other variables remain constant.
b3 = –.378; for each addition foot from the lake the price on average decreases by .378 thousand
dollars provided that the other variables remain constant.
g H 0: β i = 0
H 1: β i ≠ 0
593
Lot size: t = 1.25, p-value = .2156
Trees: t = 2.96, p-value = .0045
Distance: t = –1.94, p-value = .0577
Only for the number of trees is there enough evidence to infer a linear relationship with price.
h
A B C D
2
3 Price
4
6
8 Lower limit 35.50
10
12 Lower limit 91.86
We predict that the lot in question will sell for between $35,500 and $172,240
i
A B C D
2
3 Price
4
6
8 Lower limit -7.18
10
Estimated average price lies between $39,290 and $90,300.
594
17.2
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8734
5 R Square 0.7629
8 Observations 30
9
10 ANOVA
12 Regression 2 1223.2 611.59 43.43 0.0000
13 Residual 27 380.2 14.08
14 Total 29 1603.4
15
17 Intercept 13.01 3.53 3.69 0.0010
18 Assignment 0.194 0.200 0.97 0.3417
19 Midterm 1.11 0.122 9.12 0.0000
a ŷ = 13.01 + .194x1 + 1.11x 2
b The standard error of estimate is s ε = 3.75. It is an estimate of the standard deviation of the error
variable.
c The coefficient of determination is R 2 = .7629; 76.29% of the variation in final exam marks is
explained by the model.
d H 0: β 1 = β 2 = 0
F = 43.43, p-value = 0. There is enough evidence to conclude that the model is valid.
e b1 = .194; for each addition mark on assignments the final exam mark on average increases by
.194 provided that the other variable remains constant.

b 2 = 1.112; for each addition midterm mark the final exam mark on average increases by 1.112
provided that the other variable remains constant.
f H 0 : β1 = 0
H1 : β1 ≠ 0
t = .97, p-value = .3417. There is not enough evidence to infer that assignment marks and final
exam marks are linearly related.
g H0 : β2 = 0
H1 : β 2 ≠ 0
t = 9.12, p-value = 0. There is sufficient evidence to infer that midterm marks and final exam
marks are linearly related.
595
h
A B C D
2
3 Final
4
5 Predicted value 31
6
8 Lower limit 23
9 Upper limit 39
10
12 Lower limit 29
13 Upper limit 33
Pat’s final exam mark is predicted to lie between 23 and 39
i Pat’s predicted final grade: LCL = 12 + 14 + 23 = 49, UCL = 12 + 14 + 39 = 65
17.3a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9453
5 R Square 0.8935
8 Observations 24
9
10 ANOVA
12 Regression 4 256,793 64,198 39.86 0.0000
13 Residual 19 30,602 1,611
14 Total 23 287,395
15
17 Intercept -111.83 134.34 -0.83 0.4155
18 Permits 4.76 0.395 12.06 0.0000
19 Mortgage 16.99 15.16 1.12 0.2764
20 A Vacancy -10.53 6.39 -1.65 0.1161
21 O Vacancy 1.31 2.79 0.47 0.6446
b The standard error of estimate is s ε = 40.13. It is an estimate of the standard deviation of the
error variable.
c The coefficient of determination is R 2 = .8935; 89.35% of the variation in monthly sales of

drywall is explained by the model.
596
d H 0: β 1 = β 2 = β 3 = β 4 = 0
e b1 = 4.76; for each addition building permit monthly sales on average increase by 4.76 hundred
sheets provided that the other variables remain constant.

b 2 = 16.99; for each addition one point increase in mortgage rates monthly sales on average
increase by 16.99 hundred sheets provided that the other variables remain constant.
b3 = –10.53; for each one percentage point increase in the apartment vacancy rate monthly sales
decrease on average by 10.53 hundred sheets provided that the other variables remain constant.
b 4 = 1.31; for each one percentage point increase in the office vacancy rate monthly sales increase
on average by 1.31 hundred sheets provided that the other variables remain constant.
f H 0 : βi = 0
H1 : β i ≠ 0
Permits: t = 12.06, p-value = 0
Mortgage: t = 1.12, p-value = .2764
A Vacancy: t = –1.65, p-value = .1161
O Vacancy: t = .47, p-value = .6446
Only the number of building permits is linearly related to monthly sales.
g
A B
2
3 Drywall
4
6
8 Lower limit 167.1
9 Upper limit 352.9
10
Next month’s drywall sales are predicted to lie between 16,710 and 35,290 sheets.
597
17.4a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.5926
5 R Square 0.3511
8 Observations 126
9
10 ANOVA
12 Regression 3 3228 1075.87 22.01 0.0000
13 Residual 122 5965 48.89
14 Total 125 9192
15
17 Intercept -1.97 9.55 -0.21 0.8369
18 Minor HR 0.666 0.087 7.64 0.0000
19 Age 0.136 0.524 0.26 0.7961
20 Years Pro 1.18 0.671 1.75 0.0819
b b1 = .666; for each additional minor league home run the number of major league home runs
increases on average by .666 provided that the other variables remain constant.
b 2 = .136; for each additional year of age the number of major league home runs increases on
average by .14 provided that the other variables remain constant.
b3 = 1.18; for each additional year as a professional the number of major league home runs
increases on average by 1.18 provided that the other variables remain constant.
c s ε = 6.99 and R 2 = .3511; the model's fit is not very good.
d H 0: β 1 = β 2 = β 3 = 0
e H 0 : βi = 0
H1 : β i ≠ 0
Minor league home runs: t = 7.64, p-value = 0
Age: t = .26, p-value = .7961
Years professional: t = 1.75, p-value = .0819
At the 5% significance level only the number of minor league home runs is linearly related to the
number of major league home runs.
598
f
A B C D
2
3 Major HR
4
6
8 Lower limit 9.86
9 Upper limit 38.76
10
We predict that the player will hit between 9.86 (rounded to 10) and 38.76 (rounded to 39) home
runs.
g
A B C D
2
3 Major HR
4
6
8 Lower limit 4.88
9 Upper limit 34.25
10
It is estimated that the average player will hit between 14.66 and 24.47 home runs.
599
17.5
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8378
5 R Square 0.7020
8 Observations 50
9
10 ANOVA
12 Regression 3 399.86 133.29 36.12 0.0000
13 Residual 46 169.76 3.69
14 Total 49 569.62
15
17 Intercept 6.06 2.60 2.33 0.0244
18 Age -0.0078 0.0664 -0.12 0.9069
19 Years 0.603 0.0966 6.25 0.0000
20 Pay -0.0702 0.0524 -1.34 0.1864
a The regression equation is ŷ = 6.06 –.0078x 1 + .603x 2 –.0702x 3
b s ε = 1.92, R 2 = .7020, F = 36.12, p-value = 0. The model is valid and the fit is reasonably good.
c H 0 : βi = 0
H1 : β i ≠ 0
Age: t = –.12, p-value = .9069
Years: t = 6.25, p-value = 0
Pay: t = –1.34, p-value = .1864
Only the number of years with the company is linearly related to severance pay
d.
A B C D
2
3 Weeks SP
4
6
8 Lower limit 5.64
9 Upper limit 13.50
10
12 Lower limit 8.86
600
95% prediction interval: Lower prediction limit = 5.64, upper prediction limit = 13.50. The offer
of 5 weeks severance pay falls below the prediction interval and thus Bill is correct.
17.6
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.5369
5 R Square 0.2882
8 Observations 100
9
10 ANOVA
12 Regression 3 160.23 53.41 12.96 0.0000
13 Residual 96 395.70 4.12
14 Total 99 555.93
15
17 Intercept 0.721 1.87 0.39 0.7006
18 HS GPA 0.611 0.101 6.06 0.0000
19 SAT 0.00135 0.00144 0.94 0.3485
20 Activities 0.0462 0.0641 0.72 0.4720
b The coefficient of determination is R 2 = .2882; 28.82% of the variation in university GPAs is

explained by the model.
c H 0: β 1 = β 2 = β 3 = 0
d H 0 : βi = 0
H1 : β i ≠ 0
High school GPA: t = 6.06,, p-value = 0
SAT: t = .94, p-value = .3485
Activities: t = .72, p-value = .4720
At the 5% significance level only the high school GPA is linearly related to the university GPA
601
e
A B C D
2
3 Univ GPA
4
6
8 Lower limit 4.45
9 Upper limit 12.65
10
12 Lower limit 7.79
13 Upper limit 9.31
We predict that the student's GPA will fall between 4.45 and 12.00 (12 is the maximum).
f
A B C D
2
3 Univ GPA
4
6
8 Lower limit 4.12
9 Upper limit 10.99
10
12 Lower limit 6.90
13 Upper limit 8.22
The mean GPA is estimated to lie between 6.90 and 8.22.
602
17.7
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.4419
5 R Square 0.1953
8 Observations 25
9
10 ANOVA
12 Regression 3 34.10 11.37 1.70 0.1979
13 Residual 21 140.56 6.69
14 Total 24 174.66
15
17 Intercept 12.31 4.70 2.62 0.0160
18 Direct 0.570 1.72 0.33 0.7437
19 Newspaper 3.32 1.54 2.16 0.0427
20 Television 0.732 1.96 0.37 0.7123
a The regression equation is ŷ = 12.31 + .570x 1 + 3.32x 2 + .732x 3
b The coefficient of determination is R 2 = .1953; 19.53% of the variation in sales is explained by

the model. The coefficient of determination adjusted for degrees of freedom is .0803. The model
fits poorly.
c The standard error of estimate is s ε = 2.59. It is an estimate of the standard deviation of the error
variable
d H 0: β 1 = β 2 = β 3 = 0
F = 1.70, p-value = .1979. There is not enough evidence to conclude that the model is valid.
e H 0 : βi = 0
H1 : β i ≠ 0
Direct: t = .33, p-value = .7437
Newspaper: t = 2.16, p-value = .0427
Television: t = .37, p-value = .7123
Only expenditures on newspaper advertising is linearly related to sales.
603
f&g
A B C D
2
3 Sales
4
6
8 Lower limit 12.27
9 Upper limit 24.15
10
f We predict that sales will fall between $12,270 and $24,150.
g We estimate that mean sales will fall between $15,700 and $20,730.
h The interval in part f predicts one week’s gross sales, whereas the interval in part h estimates the
mean weekly gross sales.
17.8 a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.4125
5 R Square 0.1702
8 Observations 440
9
10 ANOVA
12 Regression 3 1205.93 401.98 29.80 1.52E-17
13 Residual 436 5880.39 13.49
14 Total 439 7086.32
15
17 Intercept 7.19 1.09 6.59 1.30E-10
18 HSize 0.0019 0.0006 3.21 0.0014
19 Children 1.10 0.14 7.84 3.58E-14
20 Adults 1.04 0.23 4.48 9.58E-06
b H 0: β 1 = β 2 = β 3 = 0
604
c b1= .0019; for each additional square foot the amount of garbage increases on average by .0019
pounds holding the other variables constant.
b2 = 1.10; for each additional child in the home the amount of garbage increases on average by
1.10 pounds holding the other variables constant.
b3= 1.04; for each additional adult at home during the day the amount of garbage increases on
average by 1.04 holding the other variables constant.
d H 0 : βi = 0
H1 : β i ≠ 0
House size : t = 3.21, p-value = .0014
Number of children: t = 7.84 p-value = 0
Number of adults at home: t = 4.48, p-value = 0
All three independent variable are linearly related to the amount of garbage.
17.9a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.5975
5 R Square 0.3570
8 Observations 40
9
10 ANOVA
12 Regression 3 1,193 397.58 6.66 0.0011
13 Residual 36 2,148 59.67
14 Total 39 3,341
15
17 Intercept 35.68 7.28 4.90 0.0000
18 Math Degree 0.247 0.070 3.54 0.0011
19 Age 0.245 0.185 1.32 0.1945
20 Income 0.133 0.153 0.87 0.3889
a ŷ = 35.68 + .247x 1 + .245x 2 + .133x 3
b H 0: β 1 = β 2 = β 3 = 0
c b1 = .247; for each one percentage point increase in the proportion of teachers with mathematics
degrees the test score increases on average by .247 provided the other variables are constant.
605
b 2 = .245; for each one year increase in mean age test score increases on average by .245 provided
the other variables are constant .
b3 = .135; for each one thousand dollar increase in salary test score increases on average by .135
provided the other variables are constant.
H 0 : βi = 0
H1 : β i ≠ 0
Proportion of teachers with at least one mathematics degree: t = 3.54, p-value = .0011
Age: t = 1.32, p-value = .1945
Income: t = .87, p-value = .3889.
The proportion of teachers with at least one mathematics degree is linearly related to test scores.
The other two variables appear to be unrelated.
d
A B C D
2
3 Test Score
4
6
8 Lower limit 49.02
9 Upper limit 81.02
10
The school's test score is predicted to fall between 49.02 and 81.02.
606
17.10a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8608
5 R Square 0.7411
8 Observations 100
9
10 ANOVA
12 Regression 4 1930 482.38 67.97 0.0000
13 Residual 95 674 7.10
14 Total 99 2604
15
17 Intercept 3.24 5.42 0.60 0.5512
18 Mother 0.451 0.0545 8.27 0.0000
19 Father 0.411 0.0498 8.26 0.0000
20 Gmothers 0.0166 0.0661 0.25 0.8028
21 Gfathers 0.0869 0.0657 1.32 0.1890
b H 0: β 1 = β 2 = β 3 = β 4 = 0
c b1 = .451; for each one year increase in the mother's age the customer's age increases on average
by .451 provided the other variables are constant (which may not be possible because of the
multicollinearity).
b 2 = .411; for each one year increase in the father's age the customer's age increases on average by
.411 provided the other variables are constant.
b3 = .0166; for each one year increase in the grandmothers' mean age the customer's age increases
on average by .0166 provided the other variables are constant.
b 4 = .0869; for each one year increase in the grandfathers' mean age the customer's age increases
on average by .0869 provided the other variables are constant.
H 0 : βi = 0
H1 : β i ≠ 0
Mothers: t = 8.27, p-value = 0
Fathers: t = 8.26, p-value = 0
Grandmothers: t = .25, p-value .8028
Grandfathers: t = 1.32, p-value = .1890
607
The ages of mothers and fathers are linearly related to the ages of their children. The other two
variables are not.
d
A B C D
2
3 Longvity
4
6
8 Lower limit 65.54
9 Upper limit 77.31
10
The man is predicted to live to an age between 65.54 and 77.31
e
A B C D
2
3 Longvity
4
6
8 Lower limit 65.65
9 Upper limit 77.77
10
The mean longevity is estimated to fall between 68.75 and 74.66.
608
17.11
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8491
5 R Square 0.7209
8 Observations 50
9
10 ANOVA
12 Regression 2 5,963 2,982 60.70 0.0000
13 Residual 47 2,309 49.1226
14 Total 49 8,272
15
17 Intercept 47.77 7.63 6.26 0.0000
18 Evaluation 0.776 1.30 0.60 0.5529
19 Articles 1.06 0.131 8.08 0.0000
Assessing the Model:
s ε = 7.01 and R 2 = .7209; the model fits well.

Testing the validity of the model:
H 0: β 1 = β 2 = 0
Drawing inferences about the independent variables:
H 0 : βi = 0
H1 : β i ≠ 0
Evaluations: t = .60, p-value = .5529
Articles: t = 8.08, p-value = 0.
The number of articles a professor publishes is linearly related to salary. Teaching evaluations are
not.
609
17.12
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8984
5 R Square 0.8072
8 Observations 50
9
10 ANOVA
12 Regression 2 9,832 4,916 98.37 0.0000
13 Residual 47 2,349 49.97
14 Total 49 12,181
15
17 Intercept -28.43 6.89 -4.13 0.0001
18 Boxes 0.604 0.0557 10.85 0.0000
19 Weight 0.374 0.0847 4.42 0.0001
a ŷ = –28.43 + .604x 1 + .374x 2
b s ε = 7.07 and R 2 = .8072; the model fits well.
c b1 = .604; for each one additional box, the amount of time to unload increases on average by
.604 minutes provided the weight is constant.

b 2 = .374; for each additional hundred pounds the amount of time to unload increases on average
by .374 minutes provided the number of boxes is constant.
H 0 : βi = 0
H1 : β i ≠ 0
Boxes: t = 10.85, p-value = 0
Weight: t = 4.42, p-value = .0001
Both variables are linearly related to time to unload.
610
d&e
A B C D
2
3 Time
4
6
8 Lower limit 35.16
9 Upper limit 66.24
10
d It is predicted that the truck will be unloaded in a time between 35.16 and 66.24 minutes.
e The mean time to unload the trucks is estimated to lie between 44.43 and 56.96 minutes.
17.13
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6584
5 R Square 0.4335
8 Observations 100
9
10 ANOVA
12 Regression 4 615.4 153.9 18.17 0.0000
13 Residual 95 804.3 8.47
14 Total 99 1419.8
15
17 Intercept 11.91 1.79 6.67 0.0000
18 Education -0.430 0.132 -3.26 0.0016
19 Age 0.0292 0.0252 1.16 0.2501
20 Children 0.0934 0.2243 0.42 0.6780
21 Income -0.0745 0.0277 -2.69 0.0085
b H 0: β 1 = β 2 = β 3 = β 4 = 0
611
e H 0 : βi = 0
H1 : β i < 0 (for beliefs 1 and 4)
H1 : β i > 0 for beliefs 2 and 3)
Belief 1: t = –3.26, p-value = .0016/2 = .0008
Belief 2: t = 1.16, p-value = .2501/2 = .1251
Belief 3: t = .42, p-value = .6780/2 = .3390
Belief 4: t = –2.69, p-value = .0085/2 = .0043
There is enough evidence to support beliefs 1 and 4. There is not enough evidence to support
beliefs 2 and 3.
17.14 a
b H 0: β 1 = β 2 = β 3 = 0
F = 24.48, p-value = 1.64E-11 .≈ 0. There is sufficient evidence to conclude that the model is
valid.
c H 0 : βi = 0
H1 : β i ≠ 0
Undergraduate GPA: t = .524, p-value = .6017
GMAT: t = 8.16, p-value = 2.71E-12 ≈ 0
612
Work experience: t = 3.00, p-value = .0036
Both the GMAT and work experience are linearly related to MBA GPA
7.15
b. H 0: β 1 = β 2 = β 3 = β 4 = β 5 = β 6 = 0
F = 17.14, p-value = 3.03E-13 ≈ 0. There is enough evidence to conclude that the model is valid.
c. H 0 : βi = 0
H1 : β i ≠ 0
Variable t-statistic p-value
Number -6.07 0
Nearest 2.60 .0108
Office space 5.80 0
Enrollment 1.59 .1159
Income 2.96 .0039
Distance -1.26 .2107
613
Intercept
The intercept is b 0 = 38.14. This is the average operating margin when all the independent
variables are zero. As we observed in Chapter 16, it is often misleading to try to interpret this
value, particularly if 0 is outside the range of the values of the independent variables (as is the case
here).
Number of Motel and Hotel Rooms

The relationship between operating margin and the number of motel and hotel rooms within 3
miles is described by b1 = −.0076. From this number we learn that in this model, for each
additional room within 3 miles of the La Quinta Inn, the operating margin decreases on average by
.0076, assuming that the other independent variables in this model are held constant. Changing the
units we can interpret b1 to say that for each additional 1,000 rooms, the margin decreases by
7.6%.
d. Distance to Nearest Competitor

The coefficient b 2 = 1.65 specifies that for each additional mile that the nearest competitor is to a
La Quinta Inn, the average operating margin increases by 1.65%, assuming the constancy of the
other independent variables.
Office Space
The relationship between office space and operating margin is expressed by b 3 = .020. Because
office space is measured in thousands of square feet, we interpret this number as the average
increase in operating margin for each additional thousand square feet of office space keeping the
other independent variables fixed. So, for every extra 100,000 square feet of office space, the
operating margin increases on average by 2.0%.
College and University Enrollment

The relationship between operating margin and college and university enrollment is described by
b 4 = .21, which we interpret to mean that for each additional thousand students the average
operating margin increases by .21% when the other variables are constant.
Median Household Income

The relationship between operating margin and median household income is described by b 5 =
.41. For each additional thousand dollar increase in median household income, the average
operating margin increases by .41%, holding all other variables constant. This statistic suggests
that motels in more affluent communities have higher operating margins.
614
Distance to Downtown Core
The last variable in the model is distance to the downtown core. Its relationship with operating
margin is described by b 6 = −.23. This tells us that for each additional mile to the downtown
center, the operating margin decreases on average by .23%, keeping the other independent
variables constant. It may be that people prefer to stay at motels that are closer to town centers.
e. & f.
e. Lower prediction limit n= 25.40, Upper prediction limit = 48.79

f. LCL = 32.97, UCL = 41.21
615
7.16
b H 0 : β1 = β2 = 0
H1 : At least one β i is not equal to zero
F = 234.93, p-value = 1.23E-88 .≈ 0. There is sufficient evidence to conclude that the model is
valid.
c H 0 : βi = 0
H1 : β i ≠ 0
PAEDUC: t = 9.73, p-value = 1.11E-11 ≈ 0
MAEDUC: t = 7.69, p-value = 2.78E-14 ≈ 0
Both variables are linearly related to EDUC.
d For each additional year of the father’s education the offspring’s education increases on average
by .219 years. For each additional year of the mother’s education the offspring’s education
increases on average by .197 years.
616
17.17
b. H 0 : β1 = β2 = β3 = β4 = 0
F = 12.62, p-value = 4.50E-10 ≈ 0. There is sufficient evidence to conclude that the model is valid.
c. H 0: β i = 0
H 1: β i ≠ 0
AGE 3.98 7.30E-05 ≈ 0
EDUC 3.221 .0013
PRESTG80 2.16 .0306
CHILDS -.73 .4631
Only CHILDS is not linearly related to EQWLTH.
d. R2= .0384; 3.84% of the variation in EQWLTH is explained by the independent variables.
617
7.18
a. H 0: β 1 = β 2 = β 3 = β 4 = β 5 = β 6 = 0
F = 9.09, p-value = 1.27E-09 ≈ 0. There is enough evidence to infer that the model is valid.
b. H 0: β i = 0
H 1: β i ≠ 0
AGE 2.34 .0194
EDUC -3.11 .0019
HRS -2.35 .0189
PRESTG80 -3.47 .0006
CHILDS -.84 .4021
EARNRS -.98 .3296
Only CHILDS and EARNRS are not linearly related to TVHOURS.
c. R2 = .0659, 6.59% of the variation in TVHOURS is explained by the independent variables.
618
7.19
a. H 0: β 1 = β 2 = β 3 = β 4 = β 5 = β 6 = 0
b. H 0: β i = 0
H 1: β i ≠ 0
AGE 4.02 6.20E-05≈ 0
EDUC 3.70 .0002
PRESTG80 .16 .8713
CHILDS .21 .8332
Only AGE and EDUC are linearly related to HELPPOOR.
c. R2 = .0295; 2.95% of the variation in HELPPOOR is explained by the independent variables.
619
17.20
a. H 0: β 1 = β 2 = β 3 = β 4 = β 5 = β 6 = β 7 = 0
b. H 0: β i = 0
H 1: β i ≠ 0
AGE .40 .6864
EDUC 7.89 9.74E-15≈ 0
HRS 7.10 2.75E-12≈ 0
CHILDS 1.61 .1084
AGEKDBRN 4.90 1.14E-06≈ 0
YEARSJOB 5.85 7.06E-09≈ 0
MOREDAYS 1.36 .1754
NUMORG 1.37 .1713
EDUC, HRS, AGEKDBRN, and YEARSJOB are linearly related to INCOME.
620
17.21
b. H 0: β 1 = β 2 = β 3 = 0
c. H 0: β i = 0
H 1: β i ≠ 0
AGE 6.93 5.99E-12 ≈ 0
EDUC 8.07 1.36E-15 ≈ 0
INCOME 1.72 .0858
AGE and EDUC are linearly related to DEFINITE.
621
17.22
b. H 0: β 1 = β 2 = β 3 = β 4 = β 5 = β 6 = 0
c. H 0: β i = 0
H 1: β i ≠ 0
DAYS1 3.33 .0009
DAYS2 .81 .4183
DAYS3 1.41 .1582
DAYS4 3.00 .0027
DAYS5 3.05 .0024
DAYS6 3.71 .0002
Only DAYS2 and DAYS3 are not linearly related to DEFINITE.
622
17.23 a
Histogram
Frequency
40
20
0
-2 -1 0 1 2 3
-3
The normality requirement is satisfied.

b
Plot of Residuals vs predicted
100
50
Residuals
0
40 60 80 100 120 140
-50
-100
Predicted
The variance of the error variable appears to be constant.
17.24
A B C D
1 Lot size Trees Distance
2 Lot size 1
3 Trees 0.2857 1
4 Distance -0.1895 0.0794 1
There does not appear to be a multicollinearity problem. The t–tests are valid.
623
17.25b
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
The normality requirement has not been violated.

c
10
8
6
4
Residuals
2
0
-2 20 30 40 50 60
-4
-6
-8
-10
Predicted
The variance of the error variable appears to be constant.

d
A B C
1 Assignment Midterm
2 Assignment 1
3 Midterm 0.1037 1
The lack of multicollinearity means that the t–tests were valid.
624
17.26a
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
The error variable may be normally distributed.

b
100
50
Residuals
0
0 100 200 300 400 500
-50
-100
Predicted

c
A B C D E
1 Permits Mortgage A Vacancy O Vacancy
2 Permits 1
3 Mortgage 0.0047 1
4 A Vacancy -0.1505 -0.0399 1
5 O Vacancy -0.1027 -0.0332 0.0652 1
Multicollinearity is not a problem.
17.27a
A B C D
1 Minor HR Age Years Pro
2 Minor HR 1
3 Age 0.0354 1
4 Years Pro -0.0392 0.7355 1
Age and years as a professional are highly correlated. The correlations of the other combinations
are small.
625
b The t–tests may not be valid.
17.28 a
Frequency Histogram
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
The error variable is approximately normally distributed.

b
2
Residuals
0
0 5 10 15 20
-2
-4
-6
Predicted

c
A B C D
1 Age Years Pay
2 Age 1
3 Years 0.8080 1
4 Pay 0.1725 0.2610 1
The correlation between age and years is high indicating that multicollinearity is a problem.
626
17.29
Histogram
Frequency
50
0
-3 -2 -1 0 1 2
Standard Residuals
5
4
3
2
Residuals
1
0
-1 3 5 7 9
-2
-3
-4
-5
Predicted
The error variable is approximately normally distributed and the variance is constant.
17.30a
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
627
6
4
2
Residuals
0
-2 15 17 19 21 23
-4
-6
Predicted
The error variable is approximately normal. However, the variance is not constant.
b
A B C D
1 Direct Newspaper Television
2 Direct 1
3 Newspaper -0.1376 1
4 Television -0.1246 0.1468 1
Multicollinearity is not a problem.
c There is one observation whose standardized residual exceeds 2.0 that should be checked
17.31
Histogram
200
Frequency
100
0
-‐3 -‐2 -‐1 0 1 2 3
Standard Residuals
The error appears to be normally distributed.
628
15
10
0
8 10 12 14 16 18 20 22
-‐5
-‐10
-‐15
There are indications that the error grows smaller as the predicted value increases. However,
overall the requirement of constant variance may be valid.
17.32 a
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
20
10
Residuals
0
-10 50 55 60 65 70 75
-20
-30
Predicted
629
c
B A C D
1 Math Degree Age Income
2 Math Degree 1
3 Age 0.0766 1
4 Income 0.0994 0.5698 1
The correlation between age and income is high. Multicollinearity is a problem.
17.33a
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3 More
Standard Residuals

b
10
8
6
4
Residuals
2
0 `
-2 60 65 70 75 80 85
-4
-6
-8
-10
Predicted

c
A B C D E
1 Mother Father Gmothers Gfathers
2 Mother 1
3 Father 0.2766 1
4 Gmothers 0.4343 0.2409 1
5 Gfathers 0.3910 0.3752 -0.0077 1
630
The correlations are large enough to cause problems with the t–tests.
17.34
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
20
10
0
40 50 60 70 80 90 100 110
-‐ 10
-‐ 20
P redic ted
The error variable appears to be normal. The error variable's variance appears to be constant. The
required conditions are satisfied.
17.35
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
631
20
10
Residuals
0
20 40 60 80 100
-10
-20
Predicted
c The error variable appears to be normally distributed. The variance of the errors appears to be
constant.
17.36a
Histogram
Frequency
40
20
0
-3 -2 -1 0 1 2 3
Standard Residuals
10
5
Residuals
0
-5 0 5 10
-5
-10
Predicted
The errors appear to be normally distributed. The variance of the errors is not constant.
632
b
Lottery Education Age Children Income
Lottery 1
Education -0.6202 1
Age 0.1767 -0.1782 1
Children -0.0230 0.1073 0.1072 1
Income -0.5891 0.7339 -0.0418 0.0801 1
There is a strong correlation between income and education. The t–tests of these two coefficients
may be distorted.
17.37
The histogram of the residuals is bell shaped.
There is no evidence of heteroscedasticity.

The requirements are satisfied.
633
17.38
The histograms is somewhat bimodal. There residuals may not be normally distributed.
There is no evidence of heteroscedasticity.
17.39 d L = 1.12, d U = 1.66. There is evidence of positive first–order autocorrelation.
17.40 d L = 1.16, d U = 1.59, 4 – d U = 2.41, 4 – d L = 2.84. There is evidence of negative first–
order autocorrelation.
17.41 d L = .95, d U = 1.89, 4 – d U = 2.11, 4 – d L = 3.05. There is evidence of first–order
autocorrelation.
17.42 d L = 1.46, d U = 1.63. There is evidence of positive first–order autocorrelation.
634
17.43 d L = 1.41, d U = 1.64, 4 – d U = 2.36, 4 – d L = 2.59. The test is inconclusive
17.44 4 – d U = 4 – 1.73 = 2.27, 4 – d L = 4 – 1.19 = 2.81. There is no evidence of negative first–
17.45 a The regression equation is ŷ = 303.3 + 14.94 x1 + 10.52 x 2
Plot of Residuals vs Time
1000
500
0
Y
0 20 40 60 80 100 120
-500
-1000
Tim e
The graph indicates that autocorrelation exists.

c
A B C
1 Durbin-Watson Statistic
2
3 d = 0.7749
d L = 1.63, d U = 1.72, 4 – d U = 2.37, 4 – d L = 2.28. There is evidence of autocorrelation.
d The model is y = β 0 + β1 x1 + β 2 x2 + β3t + ε

The regression equation is ŷ = 10.00 + 6.78 x1 + 9.37 x 2 + 9.64t
635
e
600
400
200
0
Y
0 20 40 60 80 100 120
-200
-400
-600
-800
Tim e
A B C
2
3 d = 2.1237
There is no evidence of autocorrelation.
First model: s ε = 348.7 and R 2 = .2825. Second model: s ε = 208.1 and R 2 = .7471
The second model fits better.
17.46 a The regression equation is ŷ = 2260 + .423x

b
1500
1000
500
0
-500 0 10 20 30 40 50 60
Y
-1000
-1500
-2000
-2500
Tim e
There appears to be a strong autocorrelation.
636
c
A B C
2
3 d = 0.7859
d L ≈ 1.50, d U ≈ 1.59, 4 − d U ≈ 2.41, 4 − dL ≈ ≈ 2.50. There is evidence of first–order
autocorrelation.
d The model is y = β 0 + β1 x + β 2 t + ε
The regression equation is ŷ = 446.2 + 1.10x + 38.92t
e
1000
500
0
Sales
0 10 20 30 40 50 60
-500
-1000
-1500
Tim e
A B C
2
3 d = 2.2631
There is no evidence of autocorrelation.
First model: s ε = 709.7 and R 2 = .0146. Second model: s ε = 413.7 and R 2 = .6718.
17.47
A B C
2
3 d = 1.755
d L = 1.01, d U = 1.78. There is no evidence of positive first–order autocorrelation.
637
17.48
A B C
2
3 d = 2.2003
d = 2.2003; d L = 1.30, d U = 1.46, 4 – d U = 2.70, 4 – d L = 2.54. There is no evidence of first–
17.49 a ŷ = 898.0 + 11.33x

b
Histogram
Frequency
10
5
0
-3 -2 -1 0 1 2 3
Standard Residuals
200
150
100
Residuals
50
0
-50900 950 1000 1050 1100
-100
-150
Predicted
The normality requirement is satisfied. However, the constant variance requirement is not.
A B C
2
3 d = 1.0062
d L = 1.24, d U = 1.43, 4 – d U = 2.76, 4 – d L = 2.57. There is evidence of first–order
autocorrelation.
638
c The problem is that the errors are not independent. We add a time variable (week number) to the
model. Thus, the new model is y = β 0 + β1 x + β 2 t + ε .
The regression equation is ŷ = 960.6 + 13.88x – 7.69t
Histogram
Frequency
10
5
0
-3 -2 -1 0 1 2 3
Standard Residuals
Plot of Residuals vs Years
100
50
0
Tires
800 900 1000 1100 1200

-50
-100
-150
Years
A B C
2
3 d = 1.9012
d L = 1.15, d U = 1.54, 4 – d U = 2.85, 4 – d L = 2.46. There is no evidence of first–order
autocorrelation.
d s ε = 48.55 and R 2 = .7040
Snowfall: b1 = 13.88; for each additional inch of snowfall tire sales increase on average by 13.88
(holding the time period constant).

t = 5.862, p-value = 0.
Week: b2 = −7.687; weekly sales decrease on average by 7.687 (holding snowfall constant).
t = –4.579, p-value = .0002
639
17.50a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6894
5 R Square 0.4752
8 Observations 30
9
10 ANOVA
12 Regression 2 97283 48641 12.23 0.0002
13 Residual 27 107428 3979
14 Total 29 204711
15
17 Intercept 164.01 35.9 4.57 9.60E-05
18 Fetilizer 0.140 0.081 1.72 0.0974
19 Water 0.0313 0.0067 4.64 8.08E-05
a ŷ = 164.01 + .140x 1 + .0313x 2
For each additional unit of fertilizer crop yield increases on average by .140 (holding the amount
of water constant).
For each additional unit of water crop yield increases on average by .0313 (holding the fertilizer
constant).
b H 0 : β1 = 0
H1 : β1 ≠ 0
t = 1.72, p-value = .0974. There is not enough evidence to conclude that there is a linear
relationship between crop yield and amount of fertilizer.
c H 0 : β2 = 0
H1 : β2 ≠ 0
t = 4.64, p-value = .0001. There is enough evidence to conclude that there is a linear relationship
between crop yield and amount of water.
d s ε = 63.08 and R 2 = .4752; the model fits moderately well.
640
e
Histogram
Frequency 20
10
0
-‐3 -‐2 -‐1 0 1 2 3
Standard Residuals

150
100
50
Residuals
0
0 100 200 300 400 500
-‐50
-‐100
-‐150
Predicted
The errors appear to be normal, but the plot of residuals vs predicted aeems to indicate a problem.
f
A B
2
3 Yield
4
6
8 Lower limit 69.2
9 Upper limit 349.3
10
14
We predict that the crop yield will fall between 69.2 and 349.3.
641
17.51
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6882
5 R Square 0.4736
7 Standard Error 2,644
8 Observations 40
9
10 ANOVA
12 Regression 4 220,130,124 55,032,531 7.87 0.0001
13 Residual 35 244,690,939 6,991,170
14 Total 39 464,821,063
15
17 Intercept 1,433 2,093 0.68 0.4980
18 Size -14.55 20.70 -0.70 0.4866
19 Apartments 113.0 24.01 4.70 0.0000
20 Age -50.10 98.81 -0.51 0.6153
21 Floors -223.8 171.1 -1.31 0.1994
b H 0 : β1 = β2 = β3 = β4 = 0
c The regression equation for Exercise 16.12 is ŷ = 4040 + 44.97x. The addition of the new
variables changes the coefficients of the regression line in Exercise 17.12.
642
17.52a ŷ = 29.60 − .309x 1 − 1.11x 2
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7825
5 R Square 0.6123
8 Observations 30
9
10 ANOVA
12 Regression 2 199.65 99.82 21.32 0.0000
13 Residual 27 126.44 4.68
14 Total 29 326.09
15
17 Intercept 29.60 2.08 14.22 0.0000
18 Vacancy -0.309 0.067 -4.58 0.0001
19 Unemployment -1.11 0.24 -4.73 0.0001
b R 2 = .6123; 61.23 % of the variation in rents is explained by the independent variables.
c H 0 : β1 = β 2 = 0
d H 0 : βi = 0
H1 : β i ≠ 0
Vacancy rate: t = –4.58, p-value = .0001
Unemployment rate: t = –4.73, p-value = .0001
Both vacancy and unemployment rates are linearly related to rents.
e
Histogram
Frequency
20
10
0
-3 -2 -1 0 1 2 3
Standard Residuals
643
4
Residuals
0
10 12 14 16 18 20 22
-2
-4
Predicted
The error is approximately normally distributed with a constant variance.

f
A B C
2
3 d = 2.0687
d L = 1.28, d U = 1.57, 4 – d U = 2.72, 4 – d L = 2.43. There is no evidence of first–order
autocorrelation.
g
A B C D
2
3 Rent
4
6
8 Lower limit 14.18
9 Upper limit 23.27
10
The city's office rent is predicted to lie between $14.18 and $23.27.
644
Case 17.1
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.1228
5 R Square 0.0151
8 Observations 2029
9
10 ANOVA
12 Regression 4 2,137 534.29 7.75 0.0000
13 Residual 2024 139,561 68.95
14 Total 2028 141,698
15
17 Intercept -1.15 2.20 -0.52 0.6012
18 SAT 0.0051 0.0013 3.96 0.0001
19 MBA 0.674 0.376 1.79 0.0730
20 Age -0.141 0.042 -3.31 0.0009
21 Tenure 0.082 0.176 0.47 0.6412
The model is valid (F = 7.75, p-value = 0) but the model does not fit well (R 2 = .0151; only 1.51%
of the variation in returns is explained by the model).
Interpreting the coefficients in this sample:

For each additional one–point increase in the SAT score, returns increase on average by .0051
provided the other variables remain constant.
The returns of mutual funds managed by MBAs are on average .674 larger than the returns of
mutual funds managed by people without an MBA
For each additional one–year increase in age of the manager , returns decrease on average by .141
For each additional one–year increase in the manager’s job tenure, returns increase on average by
.082 provided the other variables remain constant.
Testing the coefficients:

SAT: t = 3.96, p-value = .0001
MBA: t = 1.79, p-value = .0730
Age: t = –3.31, p-value = .0009
Tenure: t = .47, p-value = .6412
645
There is overwhelming evidence to infer that SAT scores of the undergraduate university and age
of the manager are linearly related to returns. There is weak evidence that MBAs and non–MBAs
have different mean returns. There is not enough evidence to conclude that job tenure is linearly
related to returns.
Case 17.2
Analysis of Betas
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.3597
5 R Square 0.1294
8 Observations 2029
9
10 ANOVA
12 Regression 4 15.15 3.79 75.20 0.0000
13 Residual 2024 101.97 0.050
14 Total 2028 117.12
15
17 Intercept 0.152 0.059 2.56 0.0107
18 SAT 0.00050 0.000035 14.55 0.0000
19 MBA 0.0366 0.0102 3.60 0.0003
20 Age 0.0088 0.0011 7.66 0.0000
21 Tenure -0.0352 0.0047 -7.42 0.0000
The model is valid (F = 75.20, p-value = 0) with R 2 = .1294; only 12.94% of the variation in betas
is explained by the model.

For each additional one–point increase in the SAT score, betas increase on average by .00050
The betas of mutual funds managed by MBAs are on average .0366 larger than the betas of mutual
funds managed by people without an MBA
For each additional one–year increase in age of the manager, betas increase on average by .0088
For each additional one–year increase in the manager’s job tenure, betas decrease on average by
646
SAT: t = 14.55, p-value = 0
MBA: t = 3.60, p-value = .0003
Age: t = 7.66, p-value = 0
Tenure: t = –7.42, p-value = 0
There is overwhelming evidence to infer that all four independent variables are linearly related to
mutual fund betas.
Analysis of MERs
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.2697
5 R Square 0.0728
8 Observations 2029
9
10 ANOVA
12 Regression 5 74.42 14.88 31.74 0.0000
13 Residual 2023 948.48 0.47
14 Total 2028 1022.90
15
17 Intercept 2.89 0.183 15.73 0.0000
18 SAT -0.00055 0.00011 -5.21 0.0000
19 MBA -0.082 0.0310 -2.65 0.0081
20 Age 0.0133 0.0035 3.80 0.0001
21 Tenure 0.0375 0.0145 2.59 0.0097
22 Log Assets -0.209 0.0229 -9.13 0.0000
The model is valid (F = 31.74, p-value = 0) with R 2 = .0728; only 7.28% of the variation in MERs
is explained by the model.

For each additional one–point increase in the SAT score, MERs decrease on average by .00055
The MERs of mutual funds managed by MBAs are on average .082 smaller than the MERs of
mutual funds managed by people without an MBA
For each additional one–year increase in age of the manager, MERs increase on average by .0133
647
For each additional one–year increase in the manager’s job tenure, MERs increase on average by
For each additional one–point increase in the log of the assets, MERs decrease on average by .209

SAT: t = –5.21, p-value = 0
MBA: t = –2.65, p-value = .0081
Age: t = 3.80, p-value = .0001
Tenure: t = 2.59, p-value = .0097
Log Assets: t = –9.13, p-value = 0
There is overwhelming evidence to infer that all five independent variables are linearly related to
mutual fund MERs.
648
Appendix 17
A17.1a z-test of p1 – p2 (case 1)
H 0: p 1 – p 2 = 0
H 1: p 1 – p 2 < 0
A B C D
2
3 Allicin Cold? Placebo Cold?
7 z Stat -6.96
z = −6.96; p-value = 0. There is enough evidence to conclude that garlic does help prevent colds?
b. Equal-variances t-test of µ1 − µ2
H0: (µ1 – µ2) = 0
H1: (µ1 – µ2) < 0
A B C
2
3 Allicin Days Placebo Days
4 Mean 6.29 8.11
9 df 87
10 t Stat -3.92
11 P(T<=t) one-tail 0.0001
13 P(T<=t) two-tail 0.0002
t = −3.92; p-value = .0001. There is enough evidence to conclude that garlic reduces the number of
days until recovery if a cold is caught
649
A17.2 t-test of µD
H 0: µ D = 0
H 1: µ D > 0
A B C
2
3 Eye-level Lower shelf
4 Mean 302.4 290.8
5 Variance 2482.2 6262.7
9 df 39
10 t Stat 1.35
11 P(T<=t) one-tail 0.0922
13 P(T<=t) two-tail 0.1845
t = 1.35; p-value = .0922. There is not enough evidence to conclude that placement of the product
at eye level significantly increases sales?
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 British American
4 Mean 6344.5 6358.3
5 Variance 5084.0 3104.6
9 df 59
10 t Stat -0.84
11 P(T<=t) one-tail 0.2010
13 P(T<=t) two-tail 0.4019
t = –.84, p-value = .2010. There is not enough evidence to conclude that British courses are
shorter than American courses.
650
A B C D E
1 Contingency Table
2
3 Group
4 Choice 1 2 TOTAL
5 1 7 19 26
6 2 8 17 25
7 3 11 14 25
8 TOTAL 26 50 76
9
10
12 df 2
13 p-value 0.4206
between choices students make and their level of intoxication.

A17.5 t-test of ρ or β1
H0 : ρ = 0
H1 : ρ < 0
A B C D
1 Correlation
2
3 Repair cost and Credit score
5 t Stat -6.11
6 df 217
t = −6.11; p-value = 0. There is enough evidence to infer that as the credit score increases the
repair cost decreases.
651
A B
2 Photography
4 Observations 283
5 LCL 0.085
6 UCL 0.162
Confidence interval estimate of the total number of American adults who participate in
photography
A17.7 z-test of p1 – p2 (case 1)
H0: (p1 – p2) = 0
H1: (p1 – p2) > 0
A B C D
2
3 Compression Compression & breaths
7 z Stat 5.6587
z = 5.66; p-value = 0. There is overwhelming evidence to infer that the survival rate is greater with
compression only than with compression and breaths.
A17.8 Ch-squared test of a contingency table

652
A B C D E
1 Contingency Table
2
3 Age category
4 Mutual fund 1 2 TOTAL
5 1 19 6 25
6 2 75 57 132
7 3 92 123 215
8 4 89 109 198
9 5 63 77 140
10 6 73 43 116
11 TOTAL 411 415 826
12
13
15 df 5
16 p-value 0.0001
χ 2 = 24.84; p-value = .0001. There is enough evidence to infer that the age of the head of the
household is related to whether he or she owns mutual funds.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Year 1995 Year 2005
4 Mean 66.37 64.67
5 Variance 64.24 64.71
9 df 935
10 t Stat 3.22
11 P(T<=t) one-tail 0.0007
13 P(T<=t) two-tail 0.0013
t = 3.22; p-value = .0007. There is enough evidence to infer that the ages of people who require
hip replacements are getting smaller?
653
A B C D E
1 Contingency Table
2
3 Weight category
4 Hip/Knee 1 2 TOTAL
5 1 9 6 15
6 2 113 60 173
7 3 184 166 350
8 4 165 272 437
9 TOTAL 471 504 975
10
11
13 df 3
14 p-value 0
χ 2 = 42.89; p-value = 0. There is enough evidence to conclude that weight and the joint needing
replacement are related.
A17.11 One-way analysis of variance and multiple comparisons

After the show
A B C D E F G
10 ANOVA
12 Between Groups 24.71 2 12.35 5.12 0.0067 3.03
13 Within Groups 567.11 235 2.41
14
15 Total 591.82 237
F = 5.12; p-value = .0067. There is enough evidence to infer that there are differences in memory
recall immediately after the show between the three groups.
A B C D E
2
3 LSD Omega
5 G1 After G2 After -0.152 0.594 0.577
6 G3 After -0.747 0.594 0.577
7 G2 After G3 After -0.595 0.596 0.577
Using Tukey’s method group 3 differs from both groups 1 and 2.
654
24 hours later
A B C D E F G
10 ANOVA
12 Between Groups 36.90 2 18.45 9.06 0.0002 3.03
13 Within Groups 478.53 235 2.04
14
15 Total 515.43 237
F = 9.06; p-value = .0002. There is enough evidence to infer that there are differences in memory
recall 24 hours after the show between the three groups.
A B C D E
2
3 LSD Omega
5 G1 24 hrs G2 24 hrs -0.026 0.546 0.530
6 G3 24 hrs -0.849 0.546 0.530
7 G2 24 hrs G3 24 hrs -0.823 0.547 0.530
Using either method group 3 differs from both groups 1 and 2.
A17.12
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8415
5 R Square 0.7081
8 Observations 100
9
10 ANOVA
12 Regression 2 10,744,454 5,372,227 117.6 0.0000
13 Residual 97 4,429,664 45,667
14 Total 99 15,174,118
15
17 Intercept 576.8 514.0 1.12 0.2646
18 Space 90.61 6.48 13.99 0.0000
19 Water 9.66 2.41 4.00 0.0001
a The regression equation is ŷ = 576.8 + 90.61x 1 + 9.66x 2
b The coefficient of determination is R 2 = .7081; 70.81% of the variation in electricity

consumption is explained by the model. The model fits reasonably well.
655
c H 0 : β1 = β 2 = 0
d&e
A B C D
2
3 Consumption
4
5 Predicted value 8175
6
8 Lower limit 7748
9 Upper limit 8601
10
12 Lower limit 8127
13 Upper limit 8222
e We predict that the house will consume between 7748 and 8601 units of electricity.
f We estimate that the average house will consume between 8127 and 8222 units of electricity.
A17.13Wages: Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
A B C
2
3 Goods wages Service wages
4 Mean 18.66 16.78
5 Variance 11.54 10.05
9 df 856
10 t Stat 8.37
11 P(T<=t) one-tail 1.12E-16
13 P(T<=t) two-tail 2.24E-16
t = 8.37; p-value = 0. There is overwhelming evidence to conclude that goods-producing firms pay
high wages than services-producing firms.
656
Benefits: Unequal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
A B C
2
3 Goods benefits Service benefits
4 Mean 9.82 6.34
8 df 567
9 t Stat 31.59
10 P(T<=t) one-tail 2.10E-127
12 P(T<=t) two-tail 4.19E-127
t = 31.59; p-value = 0. There is overwhelming evidence to conclude that goods-producing firms
provide larger benefits than services-producing firms.
A17.14 Multiple regression, test of coefficients

bi − βi
t=
s bi
The ordinary multiple regression model fit quite well. The coefficient of determination is .7042
and the p-value of the F-test is 0. However, no independent variable is linearly related to salary.
This is a clear sign of multicollinearity. Stepwise regression was used with the outcome shown
below.
The only independent variables that are linearly related to salary are assists in 1992-93 and goals
in 1992-93. It appears that players' salaries are most strongly related to the number of goals and
the number of assists in the previous season.
657
M N O P Q R S
1 Results of stepwise regression
2
3 Step 1 - Entering variable: Ast92_93
4
5 Summary measures
6 Multiple R 0.7725
7 R-Square 0.5967
8 Adj R-Square 0.5883
9 StErr of Est 380046.1250
10
11 ANOVA Table
12 Source df SS MS F p-value
13 Explained 1 10258242603399.2000 10258242603399.2000 71.0232 0.0000
14 Unexplained 48 6932882522112.0000 144435052544.0000
15
16 Regression coefficients
17 Coefficient Std Err t-value p-value
18 Constant 38325.8711 82112.9297 0.4667 0.6428
19 Ast92_93 25746.7754 3055.0806 8.4275 0.0000
20
21 Step 2 - Entering variable: Goal92_93
22
23 Summary measures Change % Change
24 Multiple R 0.8086 0.0361 %4.7
25 R-Square 0.6538 0.0571 %9.6
26 Adj R-Square 0.6390 0.0507 %8.6
27 StErr of Est 355859.9375 -24186.1875 -%6.4
28
29 ANOVA Table
30 Source df SS MS F p-value
31 Explained 2 11239219530119.2000 5619609765059.6100 44.3760 0.0000
32 Unexplained 47 5951905595392.0000 126636289263.6600
33
34 Regression coefficients
35 Coefficient Std Err t-value p-value
36 Constant 65924.4297 77524.0313 0.8504 0.3994
37 Ast92_93 14124.2783 5061.7593 2.7904 0.0076
38 Goal92_93 18523.1426 6655.2490 2.7832 0.0077
A17.15One-way analysis of variance

H 0 : µ1 = µ 2 = µ 3
Weight gain
A B C D E F G
10 ANOVA
12 Between Groups 608.6 2 304.32 10.47 3.73E-05 3.02
13 Within Groups 11337 390 29.07
14
15 Total 11946 392
F = 10.47; p-value = 0. There is enough evidence to infer that there are differences in weight gain
between the three groups.
658
A B C D E
2
3 LSD Omega
5 G 1 Weight gain G 2 weight gain 0.46 1.47 2.02
6 G 3 weight gain -3.95 2.24 2.02
7 G 2 weight gain G 3 weight gain -4.42 2.40 2.02
Group 3 (quitters) is different from groups 1 and 2.
Systolic blood pressure increase

A B C D E F G
10 ANOVA
12 Between Groups 929.6 2 464.78 1.67 0.1898 3.02
13 Within Groups 108622 390 278.52
14
15 Total 109552 392
F = 1.67; p-value = .1898. There is not enough evidence to conclude that there are differences in
systolic blood pressure increase between the three groups.
Diastolic blood pressure increase

A B C D E F G
10 ANOVA
12 Between Groups 676.5 2 338.24 4.80 0.0087 3.02
13 Within Groups 27463 390 70.42
14
15 Total 28139 392
F = 4.80; p-value = .0087. There is enough evidence to infer that there are differences in the
diastolic blood pressure increase between the three groups.
A B C D E
2
3 LSD Omega
5 G1 DBP gain G2 DBP gain -2.57 2.28 3.15
6 G3 DBP gain -2.99 3.48 3.15
7 G2 DBP gain G3 DBP gain -0.42 3.73 3.15
Using Tukey’s method no groups differ. Using the Bonferroni adjustment of the LSD method
groups 1 and 2 differ only.
659
A B C D
1 t-Estimate: Mean
2
3 Acres
4 Mean 676.1
6 LCL 664.1
7 UCL 688.2
Estimate of total farmland:

LCL = 229,373(664.1) = 152,326,609 acres
UCL = 229,373(688.2) = 157,854,499 acres
A17.17 Equal-variances t-test of µ1 − µ2

H0: (µ1 – µ2) = 0
H1: (µ1 – µ2) < 0
t = .338, p-value = .7357. There is not enough evidence to infer that households with at least one
union member differ from households with no union members with respect to the question about
the government doing more to solve the country’s problems.
660
A17.18
LCL = .2189, UCL = .2688
A17.19 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
z = 1.74, p-value = .0408. There is enough evidence to infer that people with at least one college-
level science course are more likely to answer correctly.
A17.20
LCL = 7.93, UCL = 9.10
661
A17.21 H 0 : The two variables are independent
χ2 = 9.60, p-value = .0082. There is sufficient evidence to conclude that the three groups of
conservatives differ in their support of capital punishment.
H0 : ρ = 0
H1 : ρ > 0
t = 5.54, p-value = 0. There is enough evidence to infer that older people watch more television.
A17.23 t-test of ρ
H0 : ρ = 0
H1 : ρ < 0
662
t = 3.79. There is not enough evidence to infer that people with higher incomes believe that
government should do more to solve the country’s problems. If anything there is evidence to the
contrary.
A17.24 Unequal-variances t-test of µ1 − µ2

H 0: µ 1 – µ 2 = 0
H 1: µ 1 – µ 2 < 0
t = -.290, p-value = .3863. There is not enough evidence to conclude that people with vision
problems are more likely to believe that it is the government’s responsibility to help pay for doctor
and hospital bills.
A17.25 t-test of ρ
H0 : ρ = 0
H1 : ρ > 0
663
t = 8.29, p-value = 0. There is enough evidence to infer that children of men with prestigious
occupations have prestigious occupations themselves.
A17.26 t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
t = -2.88, p-value = .0021. There is enough evidence to infer that there is a relationship between
the number of hours husbands and their wives work.
A17.27 t-test of ρ
H0 : ρ = 0
H1 : ρ > 0
664
t = 6.58, p-value = 0. There is enough evidence to infer that older people are more likely to vote.
A17.28 t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
t = -.647, p-value = .5178. There is not enough evidence to infer that education and amount of
time spent watching television news are related.
Case A17.1
Test of diameters:
Equal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
665
A B C
2
3 Metal Paclitaxel
4 Mean 3.01 3.00
5 Variance 0.0655 0.0667
9 df 1312
10 t Stat 0.58
11 P(T<=t) one-tail 0.2818
13 P(T<=t) two-tail 0.5637
t = .58, p-value = .5637. There is no evidence of a difference in vessel diameters between the two
groups of patients.
Test of lengths
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) ≠ 0
Two-tail F test: F =1.06 , p-value = .4527; use equal-variances t-test
A B C
2
3 Metal Paclitaxel
4 Mean 19.93 19.92
9 df 1312
10 t Stat 0.06
11 P(T<=t) one-tail 0.4772
13 P(T<=t) two-tail 0.9545
t = .06, p-value = .9545. There is no evidence of a difference in lesion lengths between the two
groups of patients.
666
Inadequate blood flow
z-test of p1 − p 2 (case 1)
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Metal Paclitaxel
7 z Stat 4.78
z = 4.78, p-value = 0. There is enough evidence to conclude that the paclitaxel stent results in a
lower incidence of inadequate blood flow.
Blockage reoccurrence
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Metal Paclitaxel
7 z Stat 5.78
z = 5.78, p-value = 0. There is enough evidence to conclude that the paclitaxel stent results in a
lower incidence of blockage reoccurrence.
667
Death from cardiac causes
H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Metal Paclitaxel
7 z Stat 0.46
10 P(Z<=z) two-tail 0.6458
z = .46, p-value = .3229. There is not enough evidence to conclude that the paclitaxel stent results
in a lower incidence of death from cardiac causes.
Death from blockage

H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
A B C D
2
3 Metal Paclitaxel
7 z Stat -0.31
10 P(Z<=z) two-tail 0.7554
z = –.31, p-value = 1–.3777 = .6223. There is not enough evidence to conclude that the paclitaxel
stent results in a lower incidence of death from blockage.
668
Case A17.2 a Chi-squared test of a contingency table
A B C D E F
1 Contingency Table
2
3 Age
4 Status 1 2 3 TOTAL
5 1 509 354 4 867
6 2 611 413 5 1029
7 3 614 437 5 1056
8 4 567 402 5 974
9 5 634 483 12 1129
10 TOTAL 2935 2089 31 5055
11
12
14 df 8
15 p-value 0.5496
χ 2 = 6.88, p-value = .5496. There is not enough evidence to conclude that there is a relationship
between age category and medical status of the driver.
b z-estimate of p
A B
2 Status
4 Observations 5055
5 LCL 0.5670
6 UCL 0.5942
The estimate of the proportion of drivers uninjured is between .5670 and .5942.
669
670
Chapter 18
18.1 a
18.2 a
671
b
2
18.3 a Sales = β 0 + β1 Space + β 2 Space + ε
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6378
5 R Square 0.4068
8 Observations 25
9
10 ANOVA
12 Regression 2 25,540 12,770 7.54 0.0032
13 Residual 22 37,248 1,693
14 Total 24 62,788
15
17 Intercept -108.99 97.24 -1.12 0.2744
18 Space 33.09 8.59 3.85 0.0009
19 Space-sq -0.666 0.177 -3.75 0.0011
s ε = 41.15 and R 2 = .4068. The model's fit is relatively poor.
F = 7.54, p-value = .0032. However, there is enough evidence to support the validity of the model.
18.4a First–order model: a Demand = β 0 + β1 Price+ ε

2
Second–order model: a Demand = β 0 + β1 Price + β 2 Price + ε
672
First–order model:
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9249
5 R Square 0.8553
8 Observations 20
9
10 ANOVA
12 Regression 1 18,798 18,798 106.44 0.0000
13 Residual 18 3,179 176.6
14 Total 19 21,977
15
17 Intercept 453.6 15.18 29.87 0.0000
18 Price -68.91 6.68 -10.32 0.0000
Second–order model:
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9862
5 R Square 0.9726
8 Observations 20
9
10 ANOVA
12 Regression 2 21,374 10,687 301.15 0.0000
13 Residual 17 603 35.49
14 Total 19 21,977
15
17 Intercept 766.9 37.40 20.50 0.0000
18 Price -359.1 34.19 -10.50 0.0000
19 Price-sq 64.55 7.58 8.52 0.0000
c The second order model fits better because its standard error of estimate is 5.96, whereas that of
the first–order models is 13.29
d ŷ .= 766.9 –359.1(2.95) + 64.55(2.95) 2 = 269.3
673
18.5a First–order model: a Time = β 0 + β1 Day+ ε
2
Second–order model: a Time = β 0 + β1 Day + β 2 Day + ε
b First–order model
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9222
5 R Square 0.8504
8 Observations 10
9
10 ANOVA
12 Regression 1 145.34 145.34 45.48 0.0001
13 Residual 8 25.56 3.20
14 Total 9 170.90
15
17 Intercept 41.40 1.22 33.90 0.0000
18 Day -1.33 0.197 -6.74 0.0001
F = 45.48, p-value = 0. The model is valid.
Second–order model
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9408
5 R Square 0.8852
8 Observations 10
9
10 ANOVA
12 Regression 2 151.28 75.64 26.98 0.0005
13 Residual 7 19.62 2.80
14 Total 9 170.90
15
17 Intercept 43.73 1.97 22.21 0.0000
18 Day -2.49 0.822 -3.03 0.0191
19 Dat-sq 0.106 0.073 1.46 0.1889
F = 26.98, p-value = .0005. The model is valid.
c The second–order model is only slightly better because its standard error of estimate is smaller.
674
18.6a MBA GPA= β 0 + β1 UnderGPA + β 2 GMAT + β 3 Work + β 4 UnderGPA × GMAT + ε
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6836
5 R Square 0.4674
8 Observations 89
9
10 ANOVA
12 Regression 4 45.97 11.49 18.43 0.0000
13 Residual 84 52.40 0.62
14 Total 88 98.37
15
17 Intercept -11.11 14.97 -0.74 0.4601
18 UnderGPA 1.19 1.46 0.82 0.4159
19 GMAT 0.0311 0.0255 1.22 0.2265
20 Work 0.0956 0.0312 3.06 0.0030
21 UGPA-GMAT -0.0019 0.0025 -0.78 0.4392
F = 18.43, p-value = 0; s ε = .790 and R 2 = .4674. The model is valid, but the fit is relatively poor.
c MBA example s ε = .788 and R 2 = .4635. There is little difference between the fits of the two
models.
18.7 a (Excel output shown below)

b H 0 : β1 = β 2 = β 3 = 0
H1 : At least on β i is not equal to 0
F = 80.65, p-value = 0. There is enough evidence to infer that the model is valid.
675
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9330
5 R Square 0.8705
7 Standard Error 4745
8 Observations 40
9
10 ANOVA
12 Regression 3 5,446,857,189 1,815,619,063 80.65 0.0000
13 Residual 36 810,445,659 22,512,379
14 Total 39 6,257,302,848
15
17 Intercept -82,044 48,530 -1.69 0.0996
18 Home % 98,443 97,463 1.01 0.3192
19 Visiting % 106,779 98,313 1.09 0.2846
20 Home-Visit 53,204 196,610 0.27 0.7882
18.8a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9255
5 R Square 0.8566
8 Observations 25
9
10 ANOVA
12 Regression 3 3398.7 1132.9 41.83 0.0000
13 Residual 21 568.8 27.08
14 Total 24 3967.4
15
17 Intercept 260.7 162.3 1.61 0.1230
18 Temperature -3.32 2.09 -1.59 0.1270
19 Currency -164.3 667.1 -0.25 0.8078
20 Temp-Curr 3.64 8.54 0.43 0.6741
676
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9312
5 R Square 0.8671
8 Observations 25
9
10 ANOVA
12 Regression 5 3440.3 688.1 24.80 0.0000
13 Residual 19 527.1 27.74
14 Total 24 3967.4
15
17 Intercept 274.8 283.8 0.97 0.3449
18 Temperature -1.72 6.88 -0.25 0.8053
19 Currency -828.6 888.5 -0.93 0.3627
20 Temp-sq -0.0024 0.0475 -0.05 0.9608
21 Curr-sq 2054.0 1718.5 1.20 0.2467
22 Temp-Curr -0.870 10.57 -0.08 0.9353
c Both models fit equally well. The standard errors of estimate and coefficients of determination
are quite similar.
18.9a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.3788
5 R Square 0.1435
8 Observations 100
9
10 ANOVA
12 Regression 3 40.38 13.46 5.36 0.0019
13 Residual 96 241.06 2.51
14 Total 99 281.44
15
17 Intercept -4.86 1.83 -2.66 0.0092
18 Faceoff 0.121 0.0366 3.31 0.0013
19 PM-diff 0.135 0.399 0.34 0.7360
20 Face-PM -0.0009 0.0080 -0.12 0.9086
677
b H 0 : β1 = β 2 = β 3 = 0
F = 5.36, p-value = .0019. There is enough evidence to infer that the model is valid.
c H0 : β 3 =0
H1 : β 3 ≠ 0
t = –.12, p-value = .9086. There is not enough evidence to infer that there is an interaction effect
between face–offs won and penalty minutes differential.
2
18.10a Yield = β 0 + β1 Pressure + β 2 Temperature + β 3 Pressure
2
+ β 4 Temperature + β 5 Pressure Temperature + ε
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8290
5 R Square 0.6872
8 Observations 80
9
10 ANOVA
12 Regression 5 42,657,846 8,531,569 32.52 0.0000
13 Residual 74 19,413,277 262,342
14 Total 79 62,071,123
15
17 Intercept 74462 7526 9.89 0.0000
18 Pressure 14.40 5.92 2.43 0.0174
19 Temperature -613.3 59.95 -10.23 0.0000
20 Press-sq -0.0159 0.0032 -5.04 0.0000
21 Temp-sq 1.23 0.12 9.86 0.0000
22 Press-temp 0.0381 0.0174 2.19 0.0316
c sε = 512 and R 2 = .6872. The model's fit is good.
18.11 The number of indicator variables is m – 1 = 5 – 1 = 4.
678
18.12 a I 1 = 1 if Catholic
I 1 = 0 otherwise
I 2 = 1 if Protestant
I 2 = 0 otherwise
b I 1 = 1 if 8:00 A.M. to 4:00 P.M.
I 1 = 0 otherwise
I 2 = 1 if 4:00 P.M. to midnight
I 2 = 0 otherwise
c I 1 = 1 if Jack Jones
I 1 = 0 otherwise
I 2 = 1 if Mary Brown
I 2 = 0 otherwise
I 3 = 1 if George Fosse
I 3 = 0 otherwise
18.13 a Macintosh b IBM c other
18.14 I1 = 1 if B.A.
= 0 otherwise
I 2 = 1 if B.B.A.
= 0 otherwise
I 3 = 1 if B.Sc. or B.Eng.
= 0 otherwise
679
I1: t = -1.54, p-value = .1269
I2: t = 2.93, p-value = .0043
I3: t = .166, p-value = .8684
Only I2 is statistically significant. However, this allows us to answer the question affirmatively.
18.15a
A B C D
2
3 MBA GPA
4
6
8 Lower limit 8.55
9 Upper limit 11.67
10
12 Lower limit 9.53
Prediction: MBA GPA will lie between 8.55 and 11.67
680
b
A B C D
2
3 MBA GPA
4
6
8 Lower limit 8.15
9 Upper limit 11.31
10
12 Lower limit 9.10
Prediction: MBA GPA will lie between 8.15 and 11.31
18.16a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8973
5 R Square 0.8051
8 Observations 100
9
10 ANOVA
12 Regression 5 2096.3 419.26 77.66 0.0000
13 Residual 94 507.5 5.40
14 Total 99 2603.8
15
17 Intercept 23.57 5.98 3.94 0.0002
18 Mother 0.306 0.0542 5.65 0.0000
19 Father 0.303 0.0476 6.37 0.0000
20 Gmothers 0.0316 0.0577 0.55 0.5853
21 Gfathers 0.0778 0.0573 1.36 0.1777
22 Smoker -3.72 0.669 -5.56 0.0000
b Exercise 18.10: ŷ = 3.24 + .451Mother + .411Father + .0166Gmothers + .0869Gfathers

There are large differences to all the coefficients.
c H0 : β5 = 0
H1 : β 5 ≠ 0
t = –5.56, p-value = 0. There is enough evidence to infer that smoking affects longevity.
681
18.17a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8368
5 R Square 0.7002
8 Observations 40
9
10 ANOVA
12 Regression 4 53,729,535 13,432,384 20.43 0.0000
13 Residual 35 23,007,438 657,355
14 Total 39 76,736,973
15
17 Intercept 3490 469.2 7.44 0.0000
18 Yest Att 0.369 0.078 4.73 0.0000
19 I1 1623 492.5 3.30 0.0023
20 I2 733.5 394.4 1.86 0.0713
21 I3 -765.5 484.7 -1.58 0.1232
b H 0 : β1 = β 2 = β 3 = β 4 = 0
c H 0 : βi = 0
H1 : β i ≠ 0
I 2 : t = 1.86, p-value = .0713
I 3 : t = –1.58, p-value = .1232

Weather is not a factor in attendance.
d H0 : β 2 = 0
H1 : β 2 > 0
t = 3.30, p-value = .0023/2 = .0012. There is sufficient evidence to infer that weekend attendance
is larger than weekday attendance.
682
18.18a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.5602
5 R Square 0.3138
8 Observations 60
9
10 ANOVA
12 Regression 2 887.9 443.95 13.03 0.0000
13 Residual 57 1941.7 34.06
14 Total 59 2829.6
15
17 Intercept 7.02 3.24 2.17 0.0344
18 Length 0.250 0.056 4.46 0.0000
19 Type -1.35 0.947 -1.43 0.1589
b H0 : β2 = 0
H1 : β 2 ≠ 0
t = –1.43, p-value = .1589. There is not enough evidence to infer that the type of commercial
affects memory test scores.
c Let
I1 = 1 if humorous
I1 = 0 otherwise
I 2 = 1 if musical
I 2 = 0 otherwise
See Excel output below.
d H 0 : βi = 0
H1 : β i ≠ 0
I1: t = 1.61, p-value = .1130
I2: t = 3.01, p-value = .0039
There is enough evidence to infer that there is a difference in memory test scores between
watchers of humorous and serious commercials.
e The variable type of commercial in parts (a) and (b) is nominal. It is usually meaningless to
conduct a regression analysis with such variables without converting them to indicator variables.
683
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6231
5 R Square 0.3882
8 Observations 60
9
10 ANOVA
12 Regression 3 1099 366.17 11.85 0.0000
13 Residual 56 1731 30.91
14 Total 59 2830
15
17 Intercept 2.53 2.15 1.18 0.2445
18 Length 0.223 0.054 4.10 0.0001
19 I1 2.91 1.81 1.61 0.1130
20 I2 5.50 1.83 3.01 0.0039
18.19 a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9233
5 R Square 0.8525
8 Observations 50
9
10 ANOVA
12 Regression 3 10384 3461.40 88.64 0.0000
13 Residual 46 1796 39.05
14 Total 49 12181
15
17 Intercept -41.42 7.00 -5.92 0.0000
18 Boxes 0.644 0.050 12.79 0.0000
19 Weight 0.349 0.075 4.65 0.0000
20 Codes 4.54 1.21 3.76 0.0005
b Let I1 = 1 if morning
I1 = 0 otherwise
I 2 = 1 if early afternoon
I 2 = 0 otherwise
684
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9727
5 R Square 0.9461
8 Observations 50
9
10 ANOVA
12 Regression 4 11525 2881.16 197.66 0.0000
13 Residual 45 655.9 14.58
14 Total 49 12181
15
17 Intercept -29.72 3.73 -7.97 0.0000
18 Boxes 0.618 0.031 19.99 0.0000
19 Weight 0.346 0.046 7.54 0.0000
20 I1 -6.76 1.50 -4.51 0.0000
21 I2 6.48 1.45 4.47 0.0001
c Model 1: s ε = 6.25 and R 2 = .8525.
Model 2: s ε = 3.82 and R 2 = .9461.

d H 0 : βi = 0
H1 : β i ≠ 0
I1: t = –4.51, p-value = 0. There is enough evidence to infer that the average time to unload in the
morning is different from that in the late afternoon.
I2: t = 4.47, p-value = .0001. There is enough evidence to infer that the average time to unload in
the early afternoon is different from that in the late afternoon.
18.20a Let
I1 = 1 if no scorecard
I1 = 0 otherwise
I 2 = 1 if scorecard overturned more than 10% of the time
I 2 = 0 otherwise
685
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7299
5 R Square 0.5327
8 Observations 100
9
10 ANOVA
12 Regression 3 1933 644.46 36.48 0.0000
13 Residual 96 1696 17.67
14 Total 99 3629
15
17 Intercept 4.65 2.06 2.26 0.0260
18 Loan Size 0.00012 0.00015 0.83 0.4084
19 I1 4.08 1.14 3.57 0.0006
20 I2 10.18 1.01 10.08 0.0000
c s ε = 4.20 and R 2 = .5327. The model's fit is mediocre.
d
A B C D E
1 Pct Bad Loan Size I1 I2
2 Pct Bad 1
3 Loan Size 0.1099 1
4 I1 -0.1653 -0.0346 1
5 I2 0.6835 0.0737 -0.5471 1
There is a high correlation between I1 and I 2 that may distort the t–tests.
e b1 =.00012; in this sample for each additional dollar lent the default rate increases by .00012
provided the other variables remain the same.

b 2 = 4.08; In this sample banks that don't use scorecards on average have default rates 4.08
percentage points higher than banks that overturn their scorecards less than 10% of the time.
b 3 = 10.18; In this sample banks that overturn their scorecards more than 10% of the time on
average have default rates 10.18 percentage points higher than banks that overturn their scorecards
less than 10% of the time.
686
f
A B C D
2
3 Pct Bad
4
6
8 Lower limit 1.39
9 Upper limit 18.49
10
12 Lower limit 8.08
We predict that the bank's default rate will fall between 1.39 and 18.49%.
18.21 a Let
I1 = 1 if welding machine
I1 = 0 otherwise
I 2 = 1 if lathe
I 2 = 0 otherwise
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7706
5 R Square 0.5938
8 Observations 60
9
10 ANOVA
12 Regression 3 193,271 64,424 27.29 0.0000
13 Residual 56 132,223 2,361
14 Total 59 325,494
15
17 Intercept 119.3 35.00 3.41 0.0012
18 Age 2.54 0.402 6.31 0.0000
19 I1 -11.76 19.70 -0.60 0.5531
20 I2 -199.4 30.71 -6.49 0.0000
687
b b1 = 2.54; in this sample for each additional month repair costs increase on average by $2.54
provided that the other variable remains constant.

b 2 = –11.76; in this sample welding machines cost on average $11.76 less to repair than stamping
machines for the same age of machine.
b 3 = –199.4; in this sample lathes cost on average $199.40 less to repair than stamping machines
for the same age of machine.
c H0 : β 2 =0
H1 : β 2 < 0
t = –.60, p-value .5531/2 = .2766. There is no evidence to infer that welding machines cost less to
repair than stamping machines.
18.22
a. The coefficient of determination in Exercise 16.107 was .3270. In this model the coefficient of
determination is .6385. This model is better.
688
b
Lower prediction limit = 150.5, upper prediction limit = 174.1

c
Lower prediction limit = 162.6, upper prediction limit = 186.3
d No, because the width of the prediction intervals are far too wide.
689
18.23a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7296
5 R Square 0.5323
8 Observations 100
9
10 ANOVA
12 Regression 5 593.9 118.78 21.40 3.08E-14
13 Residual 94 521.7 5.55
14 Total 99 1115.6
15
17 Intercept 10.26 1.17 8.76 8.12E-14
18 Wage -0.00020 0.000036 -5.69 1.43E-07
19 Pct PT -0.107 0.029 -3.62 0.0005
20 Pct U 0.060 0.012 4.83 5.38E-06
21 Av Shift 1.56 0.50 3.11 0.0025
22 UM Rel -2.64 0.492 -5.36 5.99E-07
b H0 : β4 = 0
H1 : β 4 ≠ 0
t = 3.11, p-value = .0025. There is enough evidence to infer that the availability of shiftwork
affects absenteeism.
c H0 : β5 = 0
H1 : β 5 < 0
t = –5.36, p-value = (5.99E-07) /2 = 3.00×10-7= virtually 0. There is enough evidence to infer that
in organizations where the union–management relationship is good absenteeism is lower.
690
18.24a.
b. t = -3.49, p-value = .0005. There is sufficient evidence to infer that people with similar
education and age who work for themselves earn lower incomes.
18.25 I1 = 1 if Democrat
I1 = 0 otherwise (i.e. Republican)
691
t = -12.14, p-value = 3.59E-31 ≈ 0. There is enough evidence to conclude that Democrats and
Republicans with similar values on the other four variables differ in their position on the question.
692
18.26 I1 = 1 if male
I1 = 0 otherwise (i.e. female)
t = .52, p-value = .6067. There is not enough evidence to conclude that men and women with the
same age, education, etc. differ in the amount of television watched.
18.27 I1 = 1 if strong
I1 = 0 otherwise (i.e. not very strong)
693
t = 6.54, p-value = 9.65E-11 ≈ 0. There is enough evidence to infer that people with the same age,
education, and income differ in how definite they intend to vote between those who consider
themselves strong versus not very strong supporters of their political parties.
694
18.28
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9737
5 R Square 0.9482
8 Observations 100
9
10 ANOVA
12 Regression 5 15,636,303,318 3,127,260,664 344.04 0.0000
13 Residual 94 854,451,113 9,089,905
14 Total 99 16,490,754,431
15
17 Intercept -5916 3141 -1.88 0.0627
18 Years 1022 48.93 20.88 0.0000
19 PhD 725.7 961.5 0.75 0.4523
20 Evaluation 3729 619.8 6.02 0.0000
21 Articles 439.1 80.7 5.44 0.0000
22 Gender 1090 632.0 1.72 0.0879
a H 0 : β1 = β 2 = β 3 = β 4 = β 5 = 0
b H0 : β5 = 0
H1 : β 5 > 0
t = 1.72, p-value = .0879/2 = .0440. There is evidence that male professors are better paid than
female professors with the same qualifications.
695
18.29
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8311
5 R Square 0.6907
8 Observations 8
9
10 ANOVA
12 Regression 2 38.47 19.24 5.58 0.0532
13 Residual 5 17.23 3.45
14 Total 7 55.70
15
17 Intercept 2.01 4.02 0.50 0.6385
18 Score 3.25 1.00 3.25 0.0227
19 Gender -0.039 1.35 -0.03 0.9782
In this case male–dominated jobs are paid on average $.039 (3.9 cents) less than female–
dominated jobs after adjusting for the value of each job.
18.30 All weights = .2

A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.7623
5 R Square 0.5812
8 Observations 8
9
10 ANOVA
12 Regression 2 32.37 16.19 3.47 0.1135
13 Residual 5 23.33 4.67
14 Total 7 55.70
15
17 Intercept 4.70 4.07 1.15 0.3011
18 Score 2.57 1.01 2.55 0.0514
19 Gender 0.26 1.56 0.16 0.8761
In this case male–dominated jobs are paid on average $.26 (26 cents) more than female–dominated
jobs after adjusting for the value of each job.
696
18.31 The strength of this approach lies in regression analysis. This statistical technique allows us
to determine whether gender is a factor in determining salaries. However, the conclusion is very
much dependent upon the subjective assignment of weights. Change the value of the weights and a
totally different conclusion is achieved.
18.32 a
b. In this model only AGE and PRESTG80 are included. EDUC was excluded because it is highly
correlated with PRESTG80.
697
18.33a
b. HRS has been excluded from this model. It may have been related to EDUC and AGE.
18.34
698
18.35
18.36
699
2
18.37a Mileage = β 0 + β1 Speed + β 2 Speed + ε
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8428
5 R Square 0.7102
8 Observations 50
9
10 ANOVA
12 Regression 2 1719.9 859.94 57.60 0.0000
13 Residual 47 701.64 14.93
14 Total 49 2421.5
15
17 Intercept 9.34 1.71 5.47 0.0000
18 Speed 0.802 0.077 10.39 0.0000
19 Speed-sq -0.0079 0.00073 -10.73 0.0000
c s ε = 3.86 and R 2 = .7102. The model fits moderately well.
18.38a Apply a first–order model with interaction.

b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8623
5 R Square 0.7436
8 Observations 60
9
10 ANOVA
12 Regression 3 260.2 86.74 54.14 0.0000
13 Residual 56 89.72 1.60
14 Total 59 349.9
15
17 Intercept 640.8 53.80 11.91 0.0000
18 Cars -64.17 5.27 -12.19 0.0000
19 Speed -10.63 0.897 -11.85 0.0000
20 Cars-Speed 1.08 0.088 12.26 0.0000
700
c: H 0 : β1 = β 2 = β 3 = 0
18.39a
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8668
5 R Square 0.7514
8 Observations 60
9
10 ANOVA
12 Regression 5 262.95 52.59 32.65 0.0000
13 Residual 54 86.99 1.61
14 Total 59 349.93
15
17 Intercept 404.5 327.0 1.24 0.2214
18 Cars -66.57 6.54 -10.19 0.0000
19 Speed -2.35 10.54 -0.22 0.8246
20 Cars-sq 0.107 0.097 1.10 0.2741
21 Speed-sq -0.070 0.085 -0.82 0.4180
22 Cars-Speed 1.08 0.096 11.21 0.0000
b F = 32.65, p-value = 0. There is enough evidence to infer that the model is valid.
18.40 a Let
I1 = 1 if ad was in newspaper
I1 = 0 otherwise
I 2 = 1 if ad was on radio
I 2 = 0 otherwise
701
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.6946
5 R Square 0.4824
8 Observations 52
9
10 ANOVA
12 Regression 3 90057 30019 14.91 0.0000
13 Residual 48 96627 2013
14 Total 51 186684
15
17 Intercept 282.6 17.46 16.19 0.0000
18 Ads 25.23 3.98 6.34 0.0000
19 I1 -23.36 15.83 -1.48 0.1467
20 I2 -46.59 16.44 -2.83 0.0067
b H 0 : β1 = β 2 = β3 = 0
c H 0 : βi = 0
H1 : β i ≠ 0
I–1: t = –1.48, p-value = .1467
I–2: t = –2.83, p-value = .0067
There is enough evidence to infer that the advertising medium makes a difference.
18.41 (See Excel output below)

b H0 : β3 = 0
H1 : β 3 < 0
t = –8.61, p-value = 0. There is enough evidence to infer that a team that fires its manager within
12 months wins less frequently than other teams.
702
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.9347
5 R Square 0.8736
8 Observations 50
9
10 ANOVA
12 Regression 3 0.107 0.0355 106.01 0.0000
13 Residual 46 0.0154 0.00034
14 Total 49 0.122
15
17 Intercept 0.357 0.0592 6.03 0.0000
18 BA -0.401 0.236 -1.70 0.0964
19 ERA 0.0764 0.00478 15.98 0.0000
20 Fired -0.0509 0.00591 -8.61 0.0000
2
18.42a Units = β 0 + β1 Years + β 2 Years + ε
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.4351
5 R Square 0.1893
8 Observations 100
9
10 ANOVA
12 Regression 2 175,291 87,646 11.32 0.0000
13 Residual 97 750,764 7,740
14 Total 99 926,056
15
17 Intercept 331.2 17.55 18.87 0.0000
18 Years 21.45 5.50 3.90 0.0002
19 Years-sq -0.848 0.325 -2.61 0.0105
c s ε = 87.98 and R 2 = .1893. The model fits poorly.
703
2
18.43a Depletion = β 0 + β1 Temperature + β 2 PH–level + β 3 PH–level + β 4 I 4 + β 5 I 5 + ε
where
I1 = 1 if mainly cloudy
I1 = 0 otherwise
I 2 = 1 if sunny
I 2 = 0 otherwise
b
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8085
5 R Square 0.6537
8 Observations 210
9
10 ANOVA
12 Regression 5 6596 1319 77.00 0.0000
13 Residual 204 3495 17.13
14 Total 209 10091
15
17 Intercept 1003 55.12 18.19 0.0000
18 Temperature 0.194 0.029 6.78 0.0000
19 PH Level -265.6 14.75 -18.01 0.0000
20 PH-sq 17.76 0.983 18.07 0.0000
21 I1 -1.07 0.700 -1.53 0.1282
22 I2 1.16 0.700 1.65 0.0997
c H 0 : β1 = β 2 = β 3 = β 4 = β 5 = 0
d H 0 : β1 = 0
H1 : β1 > 0
t = 6.78, p-value = 0. There is enough evidence to infer that higher temperatures deplete chlorine
more quickly.
e H0 : β3 = 0
H1 : β 3 > 0
t = 18.07, p-value = 0. There is enough evidence to infer that there is a quadratic relationship
between chlorine depletion and PH level.
704
f H 0 : βi = 0
H1 : βi ≠ 0
I1 : t = –1.53, p-value = .1282. There is not enough evidence to infer that chlorine depletion differs
between mainly cloudy days and partly sunny days.
I 2 : t = 1.65, p-value = .0997. There is not enough evidence to infer that chlorine depletion differs
between sunny days and partly sunny days.
Weather is not a factor in chlorine depletion.
705
706
Chapter 19
19.1 H0: The two population locations are the same

H1: The location of population 1 is different from the location of population 2
n1(n1 + n 2 + 1) 15(15 + 15 + 1)
E(T) = = = 232.5
n1n 2 (n12+ n 2 + 1) (152)(15)(15 + 15 + 1)
σT = = = 24.11
12 12
T − E(T) 250 − 232.5
a z= = = .73, p-value = 2P(Z > .73) = 2(1 – .7673) = .4654.
σT 24.11
T − E(T) 275 − 232.5

b z= = = 1.76, p-value 2P(Z > 1.76) = 2(1 – .9608) = .0784.
σT 24.11
c The value of the test statistic increases and the p-value decreases.
19.2 H 0 : The two population locations are the same

H1 : The location of population 1 is to the right of the location of population 2
n 1 (n 1 + n 2 + 1) 30(30 + 40 + 1)
E(T) = = = 1,065
n 1 n 2 (n21 + n 2 + 1) 2)(40)(30 + 40 + 1)
(30
σT = = = 84.26
12 12
T − E(T) 1,205 − 1,065
a z= = = 1.66, p-value P(Z > 1.66) = 1 – .9515 = .0485. There is enough
σT 84.26
evidence to infer that the location of population 1 is to the right of the location of population 2.
T − E(T) 1,065 − 1,065
b z= = = 0, p-value = P(Z > 0) = .5. There is not enough evidence to infer
σT 84.26
that the location of population 1 is to the right of the location of population 2.
c The value of the test statistic decreases and the p-value increases.

H1 : The location of population 1 is to the left of the location of population 2
Rejection region: T ≤ TL = 19
707
Sample 1 Rank Sample 2 Rank
75 5 90 9
60 1 72 3
73 4 103 10
66 2 82 8
81 7 78 6
T1 = 19 T2 = 36
There is enough evidence to infer that the location of population 1 is to the left of the location of
population 2.

H1 : The location of population 1 is different from the location of population 2
Rejection region: T ≥ TU = 127 or T ≤ TL = 83
Sample 1 Rank Sample 2 Rank
15 4.0 8 2.0
7 1.0 27 18.0
22 14.0 17 7.0
20 .5 25 16.0
32 20.0 20 11.5
18 9.5 16 5.0
26 17.0 21 13.0
17 7.0 17 7.0
23 15.0 10 3.0
30 19.0 18 9.5
T1 = 118 T2 = 92
There is not enough evidence to infer that the location of population 1 is different from the
location of population 2.

A B C D E
1 Wilcoxon Rank Sum Test
2
3 Rank Sum Observations
4 New 623.5 25
5 Leading 651.5 25
6 z Stat -0.2716
9 P(Z<=z) two-tail 0.7858
a z = –.27, p-value = .3929. There is not enough evidence to infer that the new beer is less highly
rated than the leading brand.
b The printout is identical to that of part a.
c All codes that preserve the order produce the same results.
708
A B C D E
2
4 Business 4004 40
5 Economy 8086 115
6 z Stat 3.6149
a z = 3.61, p-value = .0004. There is enough evidence to infer that the business and economy class
differ in their degree of satisfaction.

A B C D E
2
4 New 276.5 15
5 Aspirin 188.5 15
6 z Stat 1.825
a z = 1.83, p-value = .0340. There is enough evidence to infer that the new painkiller is more
effective than aspirin.
b The results are identical because the codes in this exercise and in Example 19.2 are ranked
identically.

Rejection region: z > zα = z.05 = 1.645
709
n1 (n1 + n 2 + 1) 82(82 + 75 + 1)
E (T ) = = = 6478
n1n 2 (n21 + n 2 + 1) 2 )(75)(82 + 75 + 1)
(82
σT = = = 284.6
12 12
T − E(T) 6,807 − 6,478
z= = = 1.16, p-value P(Z > 1.16) = 1 – .8770 = .1230. There is not
σT 284.6
enough evidence to infer that members of the Mathematics department rate nonparametric
techniques as more important than do members of other departments.

Rejection region: z < −zα = −z.05 = −1.645
n1(n1 + n 2 + 1) 30(30 + 30 + 1)
E(T) = = = 915
n1n 2 (n12+ n 2 + 1) (302)(30)(30 + 30 + 1)
σT = = = 67.6
12 12
T − E(T) 797 − 915
z= = = −1.75 , p-value P(Z < –1.75) = .0401. There is enough evidence to
σT 67.6
infer that companies that provide exercise programs should be given discounts.

n1(n1 + n 2 + 1) 125(125 + 125 + 1)

E(T) = = = 15,687.5
2 2
n1n 2 (n1 + n 2 + 1) (125)(125)(125 + 125 + 1)
σT = = = 571.7
12 12
T − E(T) 14,873 − 15,687.5
z= = = −1.42, p-value P(Z < – 1.42) =.0778. There is not enough
σT 571.7
evidence to infer that women are doing less housework today than last year.

710
n1(n1 + n 2 + 1) 100(100 + 100 + 1)
E(T) = = = 10,050
n1n 2 (n12+ n 2 + 1) (1002)(100)(100 + 100 + 1)
σT = = = 409.3
12 12
T − E(T) 10,691 − 10,050
z= = = 1.57, p-value P(Z > 1.57) = 1 – .9418 = .0582. There is not
σT 409.3
enough evidence to conclude that public support has decreased between this year and last year.

Rejection region: z < −zα / 2 = −z.025 = −1.96 or z > z α / 2 = z .025 = 1.96
n1(n1 + n 2 + 1) 50(50 + 50 + 1)
E(T) = = = 2525
n1n 2 (n12+ n 2 + 1) (502)(50)(50 + 50 + 1)
σT = = = 145.1
12 12
T − E(T) 2810 − 2525
z= = = 1.964, p-value = 2P(Z > 1.964), which is slightly less than 2P(Z >
σT 145.1
1.96) = 2(1 – .9750) = .05. There is enough evidence to infer that men and women experience
different levels of stomach upset.

n1(n1 + n 2 + 1) 15(15 + 25 + 1)
E(T) = = = 307.5
n1n 2 (n12+ n 2 + 1) (152)(25)(15 + 25 + 1)
σT = = = 35.8
12 12
T − E(T) 383.5 − 307.55
z= = = 2.12, p-value = P(Z > 2.12) = 1 – .9830 = .0170. There is
σT 35.8
enough evidence to infer that Tastee is superior.

711
n1(n1 + n 2 + 1) 20(20 + 20 + 1)
E(T) = = = 410,
n1n 2 (n12+ n 2 + 1) (202)(20)(20 + 20 + 1)
σT = = = 37.0
12 12
T − E(T) 439.5 − 410
z= = = .80, p-value = P(Z > .80) = 1 – .7881 = .2119. There is not enough
σT 37.0
evidence to infer that women perceive another woman wearing a size 6 dress as more professional
than one wearing a size 14 dress.
19.15a. H0: The two population locations are the same

H1: The location of population 1 is to the left of the location of population 2
n1(n1 + n 2 + 1) 125(125 + 125 + 1)

E(T) = = = 15,687.5
n1n 2 (n12+ n 2 + 1) 2125)(125 + 125 + 1)
(125)(
σT = = = 571.7
12 12
T − E(T) 13,078 − 15,687.5
z= = = −4.56, p-value = P(Z < –4.56) = 0. There is enough evidence
σT 571.7
to infer that changing the name of prunes to dried plums will increase the likelihood that shoppers
will buy.

Rejection region: z < −zα / 2 = −z.025 = −1.96 or z > z α / 2 = z .025 = 1.96
n1(n1 + n 2 + 1) 182(182 + 163 + 1)

E(T) = = = 31,486
n1n 2 (n12+ n 2 + 1) (1822)(163)(182 + 163 + 1)
σT = = = 924.9
12 12
T − E(T) 32,225.5 − 31,486
z= = = .80, p-value = 2P(Z > .80) = 2(1 – .7881) = .4238. There is
σT 924.9
not enough evidence to infer that the night and day shifts rate the service differently.
19.17 H0 :The two population locations are the same

H1:The location of population 1 is different from the location of population 2
712
z = -.353, p-value = .7241. There is not enough evidence to infer that men and women differ in
their frequency of reading newspapers.
19.18 H0 :The two population locations are the same

H1:The location of population 1 is different from the location of population 2
z = 2.43, p-value = .0153. There is enough evidence to infer that Democrats and Republicans
differ in their views about taxes for rich people.

713
z = -1.37, p-value = .1714. There is not enough evidence to infer that people who work for
themselves differ from those who work for someone else with respect to how they describe taxes
for those with high incomes.

H1: The location of population 1 (men) is to the left of the location of population 2
(women)
z = .635. There is not enough evidence to infer that men are more likely than women to lose their
jobs.

714
z = -.527, p-value = .5982. There is not enough evidence to infer that liberals and conservatives
differ in their interest in information about what’s going on in government and politics.

differ in the attention paid to local news on television about the campaign for president.

715
z = -.647, p-value = .5174. There is not enough evidence to conclude that men and women differ
in their interest in information about what’s going on in government and politics.

H1: The location of population 1 (voted for Obama) is to the right of the location of
population 2(voted for McCain)
z = 5.95, p-value = 0. There is sufficient evidence to conclude that those who voted for Obama are
more likely to believe that the economy has gotten worse in the year prior to the election.

Rejection region: z < −zα / 2 = −z.025 = −1.96 or z > zα / 2 = z.025 = 1.96
x − .5 n 15 − .5(45)
z= = = –2.24, p-value = 2P(Z < – 2.24) = 2(.0125) = .0250. There is enough
.5 n .5 45
evidence to infer that the population locations differ.

Rejection region: z < − zα = − z.10 = −1.28
x − .5 n 28 − .5(69)
z= = = –1.57, p-value = P(Z < –1.57) = .0582. There is enough evidence to
.5 n .5 69
infer that the location of population 1 is to the left of the location of population 2.

H1 : The location of population 1 is to the right of right of the location of population 2
716
x − .5 n 18 − .5(30)
z= = = 1.10, p-value = P(Z > 1.10) = 1 – .8643 = .1357. There is not enough
.5 n .5 30
evidence to infer that the location of population 1 is to the right of the location of population 2.

Pair A B Sign of Difference
1 5 3 +
2 3 2 +
3 4 4 0
4 2 3 –
5 3 3 0
6 4 1 +
7 3 3 0
8 5 4 +
9 4 2 +
10 3 5 –
11 4 1 +
12 5 2 +
13 4 2 +
14 5 3 +
15 3 2 +
16 2 2 0
x − .5n 10 − .5(12)
x = 10, n = 12, z = = = 2.31, p-value = P(Z > 2.31) = 1 – .9896 = .0104
.5 n .5 12
There is enough evidence to infer that the population 1 is located to the right of population 2.

Rejection region: z < −z α / 2 = −z .025 = −1.96 or z > zα / 2 = z.025 = 1.96
n (n + 1) 55(56) n(n + 1)(2n + 1) 55(56)(111)

E (T ) = = = 770 ; σT = = = 119.35
4 4 24 24
T − E(T) 660 − 770
z= = = –.92, p-value = 2P(z < –.92) = 2(.1788) = .3576. There is not enough
σT 119.35
evidence to infer that the population locations differ.

717
n(n + 1) 108(109) n(n + 1)(2n + 1) 108(109)(217)

E(T) = = = 2943 ; σT = = = 326.25
4 4 24 24
T − E(T) 3457 − 2943
z= = = 1.58, p-value = P(Z > 1.58) = 1 – .9429 = .0571. There is not
σT 326.25
enough evidence to conclude that population 1 is located to the right of the location of population
2.

Pair Sample 1 Sample 2 Difference |Difference| Ranks

1 9 5 4 4 5.5
2 12 10 2 2 3.5
3 13 11 2 2 3.5
4 8 9 –1 1 1.5
5 7 3 4 4 5.5
6 10 9 1 1 1.5
_____________________________
T + = 19.5 T − = 1.5
T = 19.5. There is enough evidence to infer that the population locations differ.

Pair Sample 1 Sample 2 Difference |Difference| Ranks
1 18.2 18.2 0 0
2 14.1 14.1 0 0
3 24.5 23.6 .9 .9 6.5
4 11.9 12.1 –.2 .2 2
5 9.5 9.5 0 0
6 12.1 11.3 .8 .8 5
7 10.9 9.7 1.2 1.2 8
8 16.7 17.6 –.9 .9 6.5
9 19.6 19.4 .2 .2 2
10 8.4 8.1 .3 .3 4
11 21.7 21.9 –.2 .2 2
12 23.4 21.6 1.8 1.8 9
_____________________________
T + = 34.5 T − = 10.5
T = 34.5. There is not enough evidence to conclude that the population locations differ.
718
A B C D E
1 Sign Test
2
3 Difference New - Leading
4
5 Positive Differences 46
6 Negative Differences 30
7 Zero Differences 24
8 z Stat 1.84
11 P(Z<=z) two-tail 0.0664
a z = 1.84, p-value = .0332. There is enough evidence to indicate that the new beer is more highly
rated than the leading brand.
19.34 H 0 : H 0 : The two population locations are the same

A B C D E
1 Sign Test
2
3 Difference Brand A - Brand B
4
8 z Stat 1.00
11 P(Z<=z) two-tail 0.3174
a z = 1.00, p-value = .1587. There is no evidence to infer that Brand A is preferred.

719
A B C D E
1 Sign Test
2
3 Difference European - American
4
8 z Stat 2.29
11 P(Z<=z) two-tail 0.0218
a z = 2.29, p-value = .0109. There is enough evidence to infer that the European car is perceived to
be more comfortable.
b The results are identical. All codes that preserve the order produce the same results.

A B C D E
1 Sign Test
2
3 Difference Sample 1 - Sample 2
4
8 z Stat -2.06
11 P(Z<=z) two-tail 0.0396
a. z = –2.06, p-value = .0396. There is enough evidence to infer that the population locations
differ.
720
b
A B C D E
1 Wilcoxon Signed Rank Sum Test
2
4
5 T+ 3726.5
6 T- 4148.5
7 Observations (for test) 125
8 z Stat -0.52
11 P(Z<=z) two-tail 0.6032
z = –.52, p-value = .6032. There is not enough evidence to infer that the population locations
differ.
c The sign test ignores the magnitudes of the paired differences whereas the Wilcoxon signed rank
sum test does not.
H1 :The location of population 1 is different from the location of population 2

A B C D E
1 Sign Test
2
4
8 z Stat -1.44
a z = –1.44, p-value .1490. There is not enough evidence to infer that the population locations
differ.
721
b
A B C D
2
4
5 T+ 304
6 T- 872
8 z Stat -2.91
11 P(Z<=z) two-tail 0.0036
z = –2.91, p-value = .0036. There is enough evidence to conclude that the population locations
differ.
c The sign test ignores the magnitudes of the paired differences whereas the Wilcoxon signed rank
sum test does not.

n(n + 1) 72(72 + 1) n(n + 1)(2n + 1) 72(72 + 1)(2[72] + 1)

E(T) = = = 1314 ; σT = = = 178.2
4 4 24 24
T − E(T) 378.5 − 1314
z= = = –5.25, p-value = P(Z < –5.25) = 0. There is enough evidence to
σT 178.2
infer that the drug is effective.

n(n + 1) 40(40 + 1) n(n + 1)(2n + 1) 40(40 + 1)(2[40] + 1)

E(T) = = = 410 ; σT = = = 74.4
4 4 24 24
T − E(T) 62 − 410
z= = = −4.68, p-value = P(z < –4.68) = 0. There is enough evidence to infer
σT 74.4
that women are doing less housework now than last year.
722
x − .5 n 60 − .5(98)
z= = = 2.22, p-value = P(Z > 2.22) = 1 – .9868 = .0132. There is enough
.5 n .5 98
evidence to conclude that concern about a gasoline shortage exceeded concern about an electricity
shortage.

Rejection region T ≤ TL = 73 or T ≥ TU = 203
T = 40.5. There is enough evidence of a difference between machines.

Rejection region: T ≤ TL = 110
T = 111. There is not enough evidence to infer that the swimming department has higher gross
sales.

x − .5 n 30 − .5(38)
z= = = 3.57, p-value = P(Z > 3.57) = 0. There is enough evidence to conclude
.5 n .5 38
that the European brand is preferred.

x − .5 n 5 − .5(20)
z= = = −2.24, p-value = P(Z < –2.24) = .0125. There is not enough evidence to
.5 n .5 20
conclude that children feel less pain.

Rejection region T ≤ TL = 90 or T ≥ TU = 235
T = 190. There is not enough evidence of a difference in salary offers between men and women
723
x − .5 n 32 − .5(53)
z= = = 1.51, p-value = P(Z > 1.51) = 1 – .9345 = .0655. There is not enough
.5 n .5 53
evidence to infer that preference should be given to students for high school 1.

n(n + 1) 39(39 + 1) n(n + 1)(2n + 1) 39(39 + 1)(2[39] + 1)

E(T) = = = 390 ; σT = = = 71.7
4 4 24 24
T − E(T) 48 − 390
z= = = −4.77, p-value = P(Z < –4.77) = 0. There is enough evidence to
σT 71.7
support the belief.

z = -7.40, p-value = 0. There is enough evidence to conclude that attention paid to national
network news about the presidential campaign differs from the attention paid to local news on
television about the presidential campaign.

724
z = 1.43, p-value = .1522. There is not enough evidence to conclude that attention paid to
newspaper articles differs from the attention paid to radio news about the presidential campaign.
19.50 H 0 : The locations of all 3 populations are the same.

H1 : At least two population locations differ.
Rejection region: H > χα2 , k −1 = χ.205, 2 = 5.99
⎡ 12 Tj2 ⎤ ⎡ 12 ⎛ 9842 15022 14302 ⎞⎤

H = ⎢
⎢⎣ n (n + 1)
∑ ⎥ − 3(n + 1) = ⎢
n j ⎥⎦
⎜
⎢⎣ 88(88 + 1) ⎜⎝ 23
+
36
+ ⎟⎥ − 3(88 + 1) = 1.56. There
29 ⎟⎠⎥⎦
is not enough evidence to conclude that the population locations differ.

Rejection region: H > χα2 , k −1 = χ.201,3 = 11.3
⎡ 12 T j2 ⎤ ⎡ 12 ⎛ 1207 2 1088 2 1310 2 1445 2 ⎞⎤

H = ⎢
⎢⎣ n (There
∑ ⎥ − 3(n + 1) = ⎢
n + 1) is notn enough
⎜
+ 1) ⎜⎝ that
+ + + ⎟⎥ − 3(100 + 1)
⎟⎥
= 3.28. j ⎥
⎦ evidence⎣⎢100
to(100
conclude 25the population
25 25
locations 25
differ.
⎠⎦

⎡ 12 Tj2 ⎤ ⎡ 12 ⎛ 37412 16102 49452 ⎞⎤

H = ⎢ ∑ ⎥ − 3(n + 1) = ⎢ ⎜ + + ⎟⎥ − 3(143 + 1) = 6.30.
⎢⎣ n (n + 1) n j ⎥⎦ ⎢⎣143(143 + 1) ⎜⎝ 47 29 67 ⎟⎠⎥⎦
There is enough evidence to conclude that the population locations differ.
725
1 Rank 2 Rank 3 Rank

27 8 37 14 19 6
33 12.5 12 1.5 12 1.5
18 4.5 17 3 33 12.5
29 10 . 22 7 41 15.5
41 15.5 30 11 28 9
52 17 18 4.5
75 18
_ T 1 = 85.5 T 2 = 36.5 T 3 = 49
⎡ 12 Tj2 ⎤ ⎡ 12 ⎛ 85.52 36.52 49 2 ⎞⎤
H = ⎢ ∑n ⎥ − 3(n + 1) = ⎢ ⎜ + + ⎟⎥ − 3(18 + 1) = 3.03, p-value =
⎢⎣ n (n + 1) 18(18 + 1) ⎜ 7 5 6 ⎟⎥
j ⎥
⎦ ⎢
⎣ ⎝ ⎠⎦
.2195. There is no evidence to conclude that at least two population locations differ.

1 Rank 2 Rank 3 Rank

25 10.5 19 2 27 12
15 1 21 4 25 10.5
20 3 23 8.5 22 6
22 6 . 22 6 29 15
23 8.5 28 13.5 28 13.5
T 1 = 29 T 2 = 34 T 3 = 57
⎡ 12 Tj2 ⎤ ⎡ 12 ⎛ 29 2 34 2 57 2 ⎞⎤
H = ⎢ ∑ ⎥ − 3(n + 1) = ⎢ ⎜ + + ⎟⎥ − 3(15 + 1) = 4.46, p-value = .1075.
⎢⎣ n (n + 1) n j ⎥⎦ ⎢⎣15(15 + 1) ⎜⎝ 5 5 5 ⎟⎠⎥⎦
There is not enough evidence to conclude that at least two population locations differ.

726
A B C D
1 Kruskal-Wallis Test
2
3 Group Rank Sum Observations
4 Printer 1 4889.5 50
5 Printer 2 5350 50
6 Printer 3 4864.5 50
7 Printer 4 4996 50
8
9 H Stat 0.899
10 df 3
11 p-value 0.8257
a H = .899, p-value = .8257. There is not enough evidence to conclude that differences exist
between the ratings of the four printings.

Rejection region: Fr > χα2 , k −1 = χ.210,3 = 6.25
Treatment
Block 1 Rank 2 Rank 3 Rank 4 Rank
1 10 2 12 3 15 4 9 1
2 8 2 10 3 11 4 6 1
3 13 2 14 3 16 4 11 1
4 9 1.5 9 1.5 12 3 13 4
5 7 1 8 2 14 4 10 3
T 1 = 8.5 T 2 = 12.5 T 3 = 19 T 4 = 10
⎡ k ⎤
12 ⎡ 12 2 ⎤
Fr = ⎢ 2
∑ T ⎥⎥ − 3b(k + 1) = ⎢⎣ (5)(4)(5) (8.5
j + 12.5 2 + 19 2 + 10 2 ⎥ − 3(5)(5) = 7.74, p-
⎢ b(k )(k + 1) j=1 ⎦
⎣ ⎦
value = .0517. There is enough evidence to infer that at least two population locations differ.

Rejection region: Fr > χ α2 , k −1 = χ.205, 2 = 5.99
727
Treatment
Block 1 Rank 2 Rank 3 Rank
1 7.3 2 6.9 1 8.4 3
2 8.2 3 7.0 1 7.3 2
3 5.7 1 6.0 2 8.1 3
4 6.1 1 6.5 2 9.1 3
5 5.9 1 6.1 2 8.0 3
T1= 8 T2=8 T 3 = 14
⎡ k ⎤
12 ⎡ 12 ⎤
Fr = ⎢ ∑ Tj2 ⎥ − 3b(k + 1) = ⎢ (8 2 + 8 2 + 14 2 ⎥ − 3(5)(4) = 4.8, p-value = .0907.
⎢ b(k )(k + 1)
⎣ j=1 ⎥
⎦ ⎣ (5)(3)(4) ⎦
There is not enough evidence to infer that at least two population locations differ.

A B C
1 Friedman Test
2
3 Group Rank Sum
4 Brand A 65
5 Brand B 65
6 Brand C 85
7 Brand D 85
8
9 Fr Stat 8.00
10 df 3
11 p-value 0.0460
a Fr = 8.00, p-value = .0460. There is enough evidence to infer that differences exist between the
ratings of the four brands of coffee.

b Printout is identical to that of part a.
c Different codes produce identical results provided the codes are in order.

728
A B C
1 Friedman Test
2
3 Group Rank Sum
4 Manager 1 21
5 Manager 2 10
6 Manager 3 24.5
7 Manager 4 24.5
8
9 Fr Stat 10.613
10 df 3
11 p-value 0.014
a Fr = 10.613, p-value = .0140. There is enough evidence to infer that differences exist between the
ratings of the four managers.

b The results are identical because the codes are in order.

Rejection region: H > χα2 , k −1 = χ .205, 2 = 5.99
⎡ 12 k
Tj2 ⎤ 12 ⎛ 767.5 2 917 2 1165 2 ⎞
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1) =
n j ⎥
⎜
75(75 + 1) ⎜⎝ 25
+
25
+
25
⎟ − 3(75 + 1) = 6.81, p-
⎟
⎠
⎣ ⎦
value = .0333. There is enough evidence to infer that there are differences in student satisfaction
between the teaching methods.

Rejection region: H > χα2 , k −1 = χ .205,3 = 7.81
⎡ 12 k
Tj2 ⎤
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1)
n j ⎥
⎣ ⎦
12 ⎛ 17,116.5 2 16,816.5 2 17,277 2 29,3912 ⎞

= ⎜ + + + ⎟ − 3(401 + 1) = 6.65, p-value = .0838.
401(401 + 1) ⎜⎝ 80 90 77 154 ⎟
⎠
There is not enough evidence to infer that there are differences between the four groups of GMAT
scores.
729
Orange Juice Brand

Judge 1 Rank 2 Rank 3 Rank 4 Rank
1 3 1.5 5 4 4 3 3 1.5
2 2 1 3 2 5 4 4 3
3 4 3 4 3 3 1 4 3
4 3 2 4 3 5 4 2 1
5 2 1 4 3.5 4 3.5 3 2
6 4 2 5 3.5 5 3.5 3 1
7 3 1.5 3 1.5 4 3.5 4 3.5
8 2 1 3 3 3 3 3 3
9 4 2.5 3 1 5 4 4 2.5
10 2 1 4 3 5 4 3 2
T1 = 16.5 T2 = 27.5 T3 = 33.5 T4 = 22.5
⎡ k ⎤
12 ⎡ 12 2 ⎤
Fr = ⎢ 2
∑ T ⎥⎥ − 3b(k + 1) = ⎢⎣ (10)(4)(5) (16.5
j + 27.5 2 + 33.5 2 + 22.5 2 ⎥ − 3(10)(5) =
⎢ b(k )(k + 1) j=1 ⎦
⎣ ⎦
9.42,
p-value = .0242. There is enough evidence to infer that differences in sensory perception exist
between the four brands of orange juice.
19.63a The randomized block experiment of the analysis of variance and the Friedman test should
be considered. The analysis of variance requires the number of pedestrians to be normally
distributed.
b H 0 : The locations of all 3 populations are the same.
⎡ k ⎤
12 ⎡ 12 2 ⎤
Fr = ⎢ 2
∑ T ⎥⎥ − 3b(k + 1) = ⎢⎣ (30)(3)(4) (46
j + 72 2 + 62 2 ⎥ − 3(30)(4) = 11.47, p-value =
⎢ b ( k )(k + 1) j=1 ⎦
⎣ ⎦
.0032. There is enough evidence to infer that there are differences in the number of people passing
between the three locations.

Rejection region: Fr > χ α2 ,k −1 = χ .205, 2 = 5.99
730
⎡ k ⎤
12 ⎡ 12 ⎤
Fr = ⎢ ∑ Tj2 ⎥ − 3b(k + 1) = ⎢ (28.5 2 + 22.5 2 + 212 ⎥ − 3(12)(4) = 2.63, p-value
⎢ b(k )(k + 1)
⎣ j=1 ⎥
⎦ ⎣ (12)(3)(4) ⎦
= .2691. There is not enough evidence to infer that there are differences in delivery times between
the three couriers.
19.65a The randomized block experimental design of the analysis of variance and the Friedman
test.
Rejection region: Fr > χα2 , k −1 = χ.205, 2 = 5.99
Newspaper
Job Advertised 1 Rank 2 Rank 3 Rank
Receptionist 14 2 17 3 12 1
Systems analyst 8 2 9 3 6 1
Junior secretary 25 3 20 1 23 2
Computer programmer 12 2 15 3 10 1
Legal secretary 7 2 10 3 5 1
Office manager 5 2 9 3 4 1
T1 = 13 T2 = 16 T3 = 7
⎡ k ⎤
12 ⎡ 12 2 ⎤
Fr = ⎢ 2
∑ T ⎥⎥ − 3b(k + 1) = ⎢⎣ (6)(3)(4) (13
j + 16 2 + 7 2 ⎥ − 3(6)(4) = 7.00, p-value =
⎢ b(k )(k + 1) j=1 ⎦
⎣ ⎦
.0302. There is enough evidence to infer that differences exist between the newspapers.

⎡ 12 k
Tj2 ⎤
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1)
n j ⎥
⎣ ⎦
12 ⎛ 2195 2 1650.5 2 2830 2 2102.5 2 ⎞

= ⎜ + + + ⎟ − 3(132 + 1)
132(132 + 1) ⎜⎝ 33 34 34 31 ⎟
⎠
= 14.04, p-value = .0029. There is enough evidence to conclude that there are differences in
grading standards between the four high schools.
731
⎡ k ⎤
12 ⎡ 12 2 ⎤
Fr = ⎢ 2
∑ T ⎥⎥ − 3b(k + 1) = ⎢⎣ (25)(4)(5) (59.5
j + 63.5 2 + 64 2 + 63 2 ⎥ − 3(25)(5) = .300,
⎢ b ( k )(k + 1) j=1 ⎦
⎣ ⎦
p-value = .9600. There is not enough evidence to infer that differences exist between the four
drugs.

Rejection region: Fr > χ α2 ,k −1 = χ .205, 2 = 5.99
⎡ k ⎤
12 ⎡ 12 ⎤
Fr = ⎢ ∑ Tj2 ⎥ − 3b(k + 1) = ⎢ (33 2 + 39.5 2 + 47.5 2 ⎥ − 3(20)(4) = 5.28, p-
⎢ b(k )(k + 1)
⎣ j=1 ⎥
⎦ ⎣ (20)(3)(4) ⎦
value = .0715. There is not enough evidence to infer that there are differences in the ratings of the
three recipes.
⎡ 12 k
Tj2 ⎤ 12 ⎛ 13850.52 14909.52 163902 ⎞
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1) =
n j ⎥
⎜
300(300 + 1) ⎜⎝ 100
+
100
+ ⎟ − 3(300 + 1)
100 ⎟⎠
⎣ ⎦
=4.32, p-value = .1151. There is not enough evidence to infer that differences exist between the
three shifts.
19.70a The one-way analysis of variance and the Kruskal-Wallis test should be considered.
732
⎡ 12 k
Tj2 ⎤ 12 ⎛ 4180 2 5262 2 5653 2 5005 2 ⎞
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1) =
n j ⎥
⎜
200(200 + 1) ⎜⎝ 50
+
50
+
50
+
50
⎟ − 3(200 + 1)
⎟
⎠
⎣ ⎦
= 6.96, p-value = .0733. There is not enough evidence to infer that differences exist between the
speeds at which the four brands perform.

Rejection region: H > χ 2α,k −1 = χ .205, 2 = 5.99
⎡ 12 k
Tj2 ⎤ 12 ⎛ 15652 1358.52 1171.52 ⎞
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1) =
n j ⎥
⎜
90(90 + 1) ⎜⎝ 30
+
30
+ ⎟ − 3(90 + 1) = 3.78, p-
30 ⎟⎠
⎣ ⎦
value = .1507. There is not enough evidence to infer that Democrat’s ratings of their chances
changed over the 3–month period.

⎡ 12 k
Tj2 ⎤
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1)
n j ⎥
⎣ ⎦
12 ⎛ 21,246 2 19,784 2 20,976 18,194 2 ⎞

= ⎜ + + + ⎟ − 3(400 + 1) = 4.34, p-value = .2269.
400(400 + 1) ⎜⎝ 100 100 100 100 ⎟
⎠
There is not enough evidence to infer that differences in believability exist between the four ads.

⎡ 12 k
Tj2 ⎤
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1)
n j ⎥
⎣ ⎦
12 ⎛ 28,304 2 21,285 2 21,796 2 20,4212 ⎞

= ⎜ + + + ⎟ − 3(428 + 1) = 4.64, p-value = .1999.
428(428 + 1) ⎜⎝ 123 109 102 94 ⎟
⎠
There is not enough evidence to infer that there are differences in support between the four levels
of students.
733
19.74 H0: The locations of all 5 populations are the same.
H1: At least two population locations differ.
Rejection region: H > χα2 , k −1 = χ .205, 4 = 9.49
⎡ 12 k
Tj2 ⎤
H = ⎢
⎢ n (n + 1)
∑
j=1
⎥ − 3(n + 1)
n j ⎥
⎣ ⎦
12 ⎛ 638.5 2 1233.5 2 1814.5 2 3159.5 2 2065 2 ⎞

= ⎜ + + + + ⎟ − 3(133 + 1) = 18.73, p-value =
133(133 + 1) ⎜⎝ 18 14 26 42 33 ⎟
⎠
.0009. There is enough evidence to infer that differences in perceived ease of use between the five
brands of scanners.

H = 3.80, p-value = .1494. There is not enough evidence to conclude that there are differences
between the races with respect to the question about parents and standard of living.

734
H = 26.39, p-value = 0. There is enough evidence to conclude that there are differences in health
between the five categories of marital status.

H = 8.05, p-value = .0179. There is enough evidence to infer that Democrats, Independents, and
Conservatives differ in how they describe taxes for those with high incomes.

735
H = 33.69, p-value = 0. There is enough evidence to infer that Democrats, Independents, and
Conservatives differ in their perception about whether the economy has gotten better or worse in
the previous year.

H = 5.24, p-value = .0729. There is not enough evidence to infer that Democrats, Independents,
and Conservatives differ in their expectation that the country as a whole will get better, stay the
same, or get worse in the following 12 months.
19.80 Rejection region: z < −z α / 2 = −z .025 = −1.96 or z > z α / 2 = z .025 = 1.96
z = rS n − 1 = (.23) 50 − 1 = 1.61, p-value = 2P(Z > 1.61) = 2(1 – .9463) = .1074 There is not
enough evidence to reject the null hypothesis.
19.81 H 0 : ρS = 0
H1 : ρ S > 0
736
Rejection region: rS > .497 rS = .15. There is not enough evidence to conclude that there is a
positive relationship between the two variables.
19.82 H 0 : ρS = 0
H1 : ρS ≠ 0
Mathematics a Economics b a2 b2 ab
4 5.5 5 6.5 30.25 42.25 35.75
2 3 2 1.5 9 2.25 4.5
5 7 3 4 49 16 28
4 5.5 5 6.5 30.25 42.25 35.75
2 3 3 4 9 16 12
2 3 3 4 9 16 12
1 1 2 1.5 1 2.25 1.5
Totals 28 28 137.5 137.0 129.5
n n n n n
∑ a i = 28 ∑ b i = 28 ∑ a i2 = 137.5 ∑ b i2 = 137.0 ∑a b i i = 129.5

i =1 i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑b
i =1
ai
i =1
i ⎥
⎥ 1 ⎡ (28)(28 ⎤
s ab = ∑
n − 1 ⎢⎢ i =1
a ibi −
n ⎥ = 7 − 1 ⎢129.5 − 7 ⎥ = 2.917
⎣ ⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ a i ⎥
⎟
⎢ n ⎜ ∑ ⎟ ⎥
(28) 2 ⎤
⎢ a i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢137.5 −
1 ⎡ 2
s a2 = ∑ ⎥ = 4.250, s a = s a = 4.250 = 2.062
n − 1 ⎢ i =1 n ⎥ 7 − 1 ⎣⎢ 7 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ b i ⎥
⎟
⎢ n ⎜ ∑ ⎟ ⎥
(28) 2 ⎤
⎢ b i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢137.0 −
1 ⎡ 2
s 2b = ∑ ⎥ = 4.167, s b = s b = 4.167 = 2.041
n − 1 ⎢ i =1 n ⎥ 7 − 1 ⎣⎢ 7 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s ab 2.917
rS = = = .6931
s b s b (2.062)(2.041)
H 0 : ρS = 0
H1 : ρ S ≠ 0
Rejection region: rS > .786 or rS < −.786
There is not enough evidence to infer a relationship between the grades in the two courses.
737
19.83 H 0 : ρS = 0
H1 : ρS ≠ 0
Number a Rating b a2 b2 ab
1 1 4 7 1 49 7
2 2 5 8 4 64 16
3 3 3 4.5 9 20.25 13.5
4 4 3 4.5 16 20.25 18
5 5 3 4.5 25 20.25 22.5
6 6 2 2 36 4 12
7 7 3 4.5 49 20.25 31.5
8 8 1 1 64 1 8
Totals 36 36 204 199 128.5
n n n n n
∑ a i = 36 ∑ b i = 36 ∑ a i2 = 204 ∑ b i2 = 199 ∑a b i i = 128.5

i =1 i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑b
i =1
ai
i =1
i ⎥
⎥ 1 ⎡ (36)(36 ⎤
s ab = ∑
n − 1 ⎢⎢ i =1
a ibi −
n ⎥ = 8 − 1 ⎢128.5 − 8 ⎥ = −4.786
⎣ ⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ a i ⎥
⎟
⎢ n ⎜ ∑⎟ ⎥
(36) 2 ⎤
⎢ a i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢204 −
1 ⎡ 2
s a2 = ∑ ⎥ = 6.000, s a = s a = 6.000 = 2.450
n − 1 ⎢ i =1 n ⎥ 8 − 1 ⎣⎢ 8 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ b i ⎥
⎟
⎢ n ⎜ ∑⎟ ⎥
(36) 2 ⎤
⎢ b i2 − ⎝ i =1 ⎠ ⎥ = 1 ⎢199 −
1 ⎡ 2
s 2b = ∑ ⎥ = 5.286, s b = s b = 5.286 = 2.299
n − 1 ⎢ i =1 n ⎥ 8 − 1 ⎣⎢ 8 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
s ab − 4.786
rS = = = −.8497
s b s b (2.450)(2.299)
H 0 : ρS = 0
H1 : ρ S ≠ 0
Rejection region: rS > .643or rS < −.643
There is enough evidence to infer a relationship between the number of commercials and the
rating.
19.84 H 0 : ρS = 0
H1 : ρS ≠ 0
738
Stock 1 a Stock 2 b a2 b2 ab
–7 4.5 6 8.5 20.25 72.25 38.25
–4 6.5 6 8.5 42.25 72.25 55.25
–7 4.5 –4 2 20.25 4 9
–3 8 9 12.5 64 156.25 100
2 10.5 3 5 110.25 25 52.5
–10 2.5 –3 3.5 6.25 12.25 8.75
–10 2.5 7 10.5 6.25 110.25 26.25
5 12 –3 3.5 144 12.25 42
1 9 4 6 81 36 54
–4 6.5 7 10.5 42.25 110.25 68.25
2 10.5 9 12.5 110.25 156.25 131.25
6 13 5 7 169 49 91
–13 1 –7 1 1 1 1
Totals 91 91 817 817 677.5
n n n n n
∑ a i = 91 ∑ b i = 91 ∑ a i2 = 817 ∑ b i2 = 817 ∑a b i i = 677.5

i =1 i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑b
i =1
ai
i =1
i ⎥
⎥ 1 ⎡ (91)(91⎤
s ab = ∑
n − 1 ⎢⎢ i =1
a ibi −
n ⎥ = 13 − 1 ⎢677.5 − 13 ⎥ = 3.375
⎣ ⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ a i ⎥
⎟
1 ⎢ n ⎜ ∑ ⎟ ⎥
1 ⎡ (91)2 ⎤
s a2 = ⎢ a i2 − ⎝ i =1 ⎠ ⎥ =
∑ ⎢817 −
2
⎥ = 15.00, sa = sa = 15.00 = 3.873
n − 1 ⎢ i =1 n ⎥ 13 − 1 ⎣⎢ 13 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ b i ⎥
⎟
1 ⎢ n ⎜ ∑ ⎟ ⎥
1 ⎡ (91)2 ⎤
s 2b = ⎢ b i2 − ⎝ i =1 ⎠ ⎥ =
∑ ⎢817 −
2
⎥ = 15.00, s b = s b = 15.00 = 3.873
n − 1 ⎢ i =1 n ⎥ 13 − 1 ⎣⎢ 13 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
sab 3.375
rS = = = .2250
s bs b (3.873)(3.873)
H 0 : ρS = 0
H1 : ρ S ≠ 0
There is not enough evidence to infer a relationship between the two stock returns.
19.85 H 0 : ρS = 0
H1 : ρS ≠ 0
739
Experience a Rating b a2 b2 ab
1 1.5 1 2 2.25 4 3
17 13 4 16 169 256 208
20 17 4 16 289 256 272
9 6.5 5 21.5 42.25 462.25 139.75
2 3 2 6 9 36 18
13 11 4 16 121 256 176
9 6.5 3 11 42.25 121 71.5
23 23 5 21.5 529 462.25 494.5
7 4 2 6 16 36 24
10 8 5 21.5 64 462.25 172
12 10 5 21.5 100 462.25 215
24 24 2 6 576 36 144
8 5 2 6 25 36 30
20 17 5 21.5 289 462.25 365.5
21 20 3 11 400 121 220
19 15 2 6 225 36 90
1 1.5 1 2 2.25 4 3
22 22 3 11 484 121 242
20 17 4 16 289 256 272
11 9 3 11 81 121 99
18 14 5 21.5 196 462.25 301
14 12 4 16 144 256 192
21 20 3 11 400 121 220
21 20 1 2 400 4 40
Totals 300 300 4,895 4,850.5 4,012.25
n n n n n
∑ a i = 300 ∑ b i = 300 ∑ a i2 = 4,895 ∑ b i2 = 4,850.5 ∑a b i i = 4,012.25

i =1 i =1 i =1 i =1 i =1
⎡ n n ⎤
1 ⎢
⎢ n ∑ ∑b
i =1
ai
i =1
i ⎥
⎥ 1 ⎡ (300)(300) ⎤
s ab = ∑
n − 1 ⎢⎢ i =1
a ibi −
n ⎥ = 24 − 1 ⎢4,012.25 −
⎣ 24 ⎥ = 11.40
⎦
⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ a i ⎥
⎟
1 ⎢ n ⎜ ∑
⎟ ⎥
1 ⎡ (300)2 ⎤
s a2 = ⎢ a i2 − ⎝ i =1 ⎠ ⎥ =
∑ ⎢4,895 −
2
⎥ = 49.78, sa = sa = 49.78 = 7.056
n − 1 ⎢ i =1 n ⎥ 24 − 1 ⎣⎢ 24 ⎦⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
⎡ 2
⎛ n ⎞ ⎤
⎢ ⎜ b i ⎟ ⎥
1 ⎢ n ⎜ ∑
⎟ ⎥
1 ⎡ (300)2 ⎤
s 2b = ⎢ b i2 − ⎝ i =1 ⎠ ⎥ =
∑ ⎢4,850.5 −
2
⎥ = 47.85, s b = s b = 47.85 = 6.917
n − 1 ⎢ i =1 n ⎥ 24 − 1 ⎢⎣ 24 ⎥⎦
⎢ ⎥
⎢ ⎥
⎣ ⎦
sab 11.40
rS = = = .2336
sbsb (7.056)(6.917)
740
There is not enough evidence to infer a relationship between experience and quality.
19.86 H 0 : ρS = 0
H1 : ρ S ≠ 0
A B C D
1 Spearman Rank Correlation
2
3 Price and Odometer
4 Spearman Rank Correlation -0.0201
5 z Stat -0.20
7 z Critical one tail 1.6449
8 P(Z<=z) two tail 0.8412
9 z Critical two tail 1.96
z = –.20, p-value = .8412. There is not enough evidence to infer that odometer reading and price
are related.
19.87 H 0 : ρS = 0
H1 : ρ S ≠ 0
A B C D
2
3 Rating and Grade
5 z Stat -2.93
z = –2.93, p-value = .0034. There is enough evidence to support the theory.
19.88 H 0 : ρS = 0
H1 : ρ S > 0
741
A B C D
2
3 Age and Heartburn
4 Spearman Rank Correlation 0.0302
5 z Stat 0.54
z = .54, p-value = .2931. There is not enough evidence to conclude that age and severity of
heartburn are positively related.
19.89 H 0 : ρS = 0
H1 : ρS > 0
A B C D
2
3 Test and Length
5 z Stat 4.19
8 P(Z<=z) two tail 0
z = 4.19, p-value = 0. There is enough evidence to conclude that the longer the commercial the
higher the test score will be.
19.90 H 0 : ρS = 0
H1 : ρS > 0
A B C D
2
3 Floor and Price
5 z Stat 3.87
z = 3.87, p-value = .0001. There is sufficient evidence to conclude that price and floor number are
positively related.
742
19.91 H 0 : ρS = 0
H1 : ρS < 0
A B C D
2
3 Cigarettes and Taste
5 z Stat -5.05
z = –5.05, p-value = 0. There is sufficient evidence to conclude that the more a person smokes the
less taste sensation he or she has.
19.92 H 0: ρ s= 0
H 1: ρ s> 0
A B C D
2
3 Wager and Enjoyment
5 z Stat 5.52
z = 5.52, p-value = 0. There is enough evidence to infer that the greater the wager the more
enjoyable the game is.
19.93 H 0: ρ s= 0
H 1: ρ s ≠ 0
743
z = 2.17, p-value = .0302. There is enough evidence to conclude that age and belief about taxes for
those with high incomes are related.
19.94 H 0: ρ s= 0
H 1: ρ s< 0
z = -2.13, p-value = .0166. There is enough evidence to conclude that more educated people are
more likely to believe that compared to their parents their standard of living is better.
19.95 H 0: ρ s= 0
H 1: ρ s < 0
z = -9.59, p-value = 0. There is enough evidence to conclude that more educated people are more
likely to believe that they are healthier.
744
19.96 H 0: ρ s= 0
H 1: ρ s < 0
z = -5.25, p-value = 0. There is enough evidence to conclude that older people are more interested
in information about government and politics.
19.97 H 0: ρ s= 0
H 1: ρ s> 0
z = 14.68, p-value = 0. There is enough evidence to conclude that people who believe that the
economy has gotten worse in the previous year also believe that unemployment has gotten worse
in the previous year.
19.98 H 0 : ρS = 0
H1 : ρS > 0
745
A B C D
2
3 Education and Income
5 z Stat 9.64
z = 9.64, p-value = 0. There is sufficient evidence to conclude that more education and higher
incomes are linked.

A B C D E
2
4 Section 1 15297.5 113
5 Section 2 14592.5 131
6 z Stat 2.65
z = 2.65, p-value = .0082. There is enough evidence to infer that the two teaching methods differ

H1 : The location of population 1 is left of the location of population 2
A B C D E
2
4 New 207.5 15
5 Existing 257.5 15
6 z Stat -1.04
z = –1.04, p-value = .1499. There is not enough evidence to infer that the new method is better.
746
A B C
1 Friedman Test
2
3 Group Rank Sum
4 Typeface 1 50.5
5 Typeface 2 38
6 Typeface 3 66
7 Typeface 4 45.5
8
9 Fr Stat 12.615
10 df 3
11 p-value 0.0055
Fr = 12.615, p-value = .0055. There is enough evidence to conclude that there are differences
between typefaces.

A B C D E
1 Sign Test
2
3 Difference Drug A - Drug B
4
8 z Stat -3.58
11 P(Z<=z) two-tail 0.0004
z = –3.58, p-value = .0002. There is enough evidence to conclude that drug B is more effective.

747
A B C D
2
3 Difference Drug A - Drug B
4
5 T+ 36
6 T- 342
8 z Stat -3.68
11 P(Z<=z) two-tail 0.0002
Note that the number of non-zero observations is less than 30, invalidating the use of z statistic. T
= 36, rejection region: T ≤ 120. There is enough evidence to conclude that drug B is more
effective.
19.104a The one-way analysis of variance and the Kruskal-Wallis test should be considered. If the
data are normal apply the analysis of variance, otherwise use the Kruskal-Wallis test.
A B C D
2
4 Binding 1 827 25
5 Binding 2 1110 25
6 Binding 3 913 25
7
8 H Stat 3.55
9 df 2
10 p-value 0.1699
H = 3.55, p-value = .1699. There is not enough evidence to infer that there are differences between
bindings.

748
A B C D E
2
4 New Material 2747 50
5 Old Material 2303 50
6 z Stat 1.53
z = 1.53, p-value = .0630. There is not enough evidence to conclude that the new material takes
longer to burst into flames.

A B C D
2
4 Sunday 10060 63
5 Monday 2977 26
6 Tuesday 2932.5 29
7 Wednesday 3834.5 31
8 Thursday 4060.5 30
9 Friday 6045 42
10 Saturday 6405.5 48
11
12 H Stat 14.87
13 df 6
14 p-value 0.0213
H = 14.87, p-value = .0213. There is enough evidence to infer that there are differences in the
perceptions of speed of service between the days of the week.

749
A B C D E F
1 Sign Test
2
3 Difference Commercial 1 - Commercial 2
4
8 z Stat -1.00
11 P(Z<=z) two-tail 0.3174
z = –1.00, p-value = .3174. There is not enough evidence to infer differences in believability
between the two commercials.

A B C D
2
3 Difference Men - Women
4
5 T+ 324
6 T- 204
8 z Stat 1.12
11 P(Z<=z) two-tail 0.2618
z = 1.12, p-value = .1309. There is not enough evidence to conclude that men lose a greater
percentage of their hearing than women.

750
A B C D E
2
4 This Year 37525.5 200
5 10 Years Ago 42674.5 200
6 z Stat -2.23
z = –2.23, p-value = .0130. There is enough evidence to infer that people perceive newspapers as
doing a better job 10 years ago than today.
19.110 H 0 : ρS = 0
H1 : ρ S ≠ 0
A B C D
2
3 Reference and GPA
5 z Stat 1.05
z = 1.05, p-value = .2960. There is not enough evidence to infer that the letter of reference and
MBA GPA are related.

A B C D E
2
4 Males 10336.5 100
5 Females 9763.5 100
6 z Stat 0.70
z = .70, p-value = .4840. There is not enough evidence to conclude that businesswomen and
business men differ in the number of business trips taken per year.
751
A B C D E
1 Sign Test
2
3 Difference Before - After
4
8 z Stat 2.86
11 P(Z<=z) two-tail 0.0042
z = 2.86, p-value = .0021. There is enough evidence to infer that the midterm test negatively
influences student opinion.

A B C D E
2
4 2 Years Ago 10786.5 100
5 ThisYear 9313.5 100
6 z Stat 1.80
z = 1.80, p-value = .0360. There is enough evidence to indicate that the citizens of Stratford
should be concerned.

752
A B C D E
2
4 Low 9055 100
5 High 11045 100
6 z Stat -2.43
z = –2.43, p-value = .0075. There is enough evidence to conclude that boys with high levels of
lead are more aggressive than boys with low levels.
19.115a H 0 : The two population locations are the same

A B C D E F G
1 Sign Test
2
3 Difference Female Professor - Male Professor
4
8 z Stat 5.27
9 P(Z<=z) one-tail 0
z = 5.27, p-value = 0. There is enough evidence to infer that female students rate female professors
higher than they rate male professors.
b H 0 : The two population locations are the same
A B C D E F G
1 Sign Test
2
3 Difference Female Professor - Male Professor
4
8 z Stat -1.39
11 P(Z<=z) two-tail 0.1656
753
z = –1.39, p-value = .0828. There is not enough evidence to infer that male students rate male
professors higher than they rate female professors.

A B C D
2
4 Unattractive 16844.5 134
5 Neutral 13313 68
6 Attractive 26122.5 133
7
8 H Stat 42.59
9 df 2
10 p-value 0
H = 42.59, p-value = 0. There is enough evidence to conclude that incomes of lawyers are affected
by physical attractiveness.

A B C D E
2
4 Telecommuters 10934.5 100
5 Office 9165.5 100
6 z Stat 2.16
z = 2.16, p-value = .0153. There is enough evidence to conclude that telecommuters are more
satisfied with their jobs.

754
A B C D E
2
4 3 Hours Before 22553.5 180
5 Closing 42426.5 180
6 z Stat -10.06
z = –10.06, p-value = 0. There is enough evidence to conclude that alcohol impairs judgment.

Day 1versus Before
A B C D E
1 Sign Test
2
3 Difference Before - Day 1
4
8 z Stat 8.19
z = 8.19, p-value = 0
Day 2 versus Before
A B C D E
1 Sign Test
2
4
8 z Stat 7.68
z = 7.68, p-value = 0
Day 3 versus Before
755
A B C D E
1 Sign Test
2
4
8 z Stat 5.92
z = 5.92, p-value = 0
There is enough evidence to infer that exercisers who abstain from physical activity are less happy
than when they are exercising.
b H 0 : The two population locations are the same

H1 : The location of population 1 (Day 2) is to the left of the location of population 2
(Day 3)
A B C D E
1 Sign Test
2
3 Difference Day 2 - Day 3
4
8 z Stat -5.38
z = –5.38, p-value = 0. There is enough evidence to conclude that by the third day their moods
were improving.
c 1. Exercisers who abstained were adjusting to their inactivity by the third day. 2. By the third
day exercisers realized that they were closer to the end of their inactivity.
Case 19.1
a Wilcoxon rank sum tests
H 0 : The two population locations are the same.
H 1 : The location of population 1 is to the left of the location of population 2.
756
Quality of work
A B C D E
2
4 No 875.5 26
5 Yes 8169.5 108
6 z Stat -4.95
z = –4.95, p-value = 0. There is overwhelming evidence to infer that customers who say they will
return assess quality of work higher than customers who do not plan to return.
Fairness of price
A B C D E
2
4 No 1477 26
5 Yes 7568 108
6 z Stat -1.56
z = = –1.56, p-value = .0589. There is not enough evidence to infer that customers who say they
will return assess fairness of price higher than customers who do not plan to return.
Explanation of work and guarantee

A B C D E
2
4 No 1656 26
5 Yes 7389 108
6 z Stat -0.56
z = –.56, p-value = .2888. There is no evidence to infer that customers who say they will return
assess explanation of work and guarantee higher than customers who do not plan to return.
757
Checkout process
A B C D E
2
4 No 1461 26
5 Yes 7584 108
6 z Stat -1.65
z = –1.65, p-value = .0490. There is evidence to infer that customers who say they will return
assess the checkout process higher than customers who do not plan to return.
b Kruskal-Wallis tests
H 0 : The location of all 3 populations is the same
H 1 : At least two population locations differ
Quality of work
A B C D
2
4 Positive 2495.5 33
5 Negative 1018 21
6 No comment 5531.5 80
7
8 H Stat 6.63
9 df 2
10 p-value 0.0364
H = 6.63, p-value = .0364. There is evidence to conclude that customers who make positive,
negative, and no comment differ in their assessment of quality of work performed.
758
Fairness of price
A B C D
2
5 Negative 1143.5 21
6 No comment 5520 80
7
8 H Stat 2.97
9 df 2
10 p-value 0.2268
H = 2.97, p-value = .2268. There is no evidence to conclude that customers who make positive,
negative, and no comment differ in their assessment of fairness of price
Explanation of work and guarantee

A B C D
2
4 Positive 2291 33
5 Negative 1335 21
6 No comment 5419 80
7
8 H Stat 0.299
9 df 2
10 p-value 0.8611
H = .299, p-value = .8611. There is no evidence to conclude that customers who make positive,
negative, and no comment differ in their assessment of explanation of work and guarantee.
Checkout process
A B C D
2
5 Negative 1200 21
6 No comment 5911.5 80
7
8 H Stat 5.40
9 df 2
10 p-value 0.0672
759
H = 5.40, p-value = .0672. There is not enough evidence to conclude that customers who make
positive, negative, and no comment differ in their assessment of the checkout process.
760
Appendix 19
A B C D
1 t-Estimate: Mean
2
3 Plastic bags
4 Mean 1495
6 LCL 1451
7 UCL 1539
Estimate of the total number of plastic bags

LCL = 112 million(1451) = 162.512 billion
UCL = 112 million(1539) = 172.368 billion
A19.2 Spearman rank correlation coefficient test
H 0: ρ S = 0
H 1: ρ S > 0
A B C D
2
3 Satisfaction and Time
5 z Stat 8.984
z = 8.98; p-value = 0. There is enough evidence to infer that those who do more research are more
satisfied with their choice.
A19.3 Sign test

H0: The location of the two populations are the same
H1: The location of population 1 is to the right of the location of population2
761
A B C D E
1 Sign Test
2
3 Difference Cork rate - Metal rate
4
8 z Stat 4.75
z = 4.75; p-value = 0. There is sufficient evidence to indicate that wine bottled with a screw cap is
perceived to be inferior.
A19.4 Histograms (not shown) are approximately bell shaped

Unequal-variances t-test of µ1 − µ 2
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 British American
4 Mean 7137 9304
5 Variance 38051 110151
8 df 53
9 t Stat -31.61
10 P(T<=t) one-tail 2.26E-36
12 P(T<=t) two-tail 4.51E-36
t = −31.61; p-value = 0. There is overwhelming evidence to conclude that the total distance of
American golf courses is greater than that of British courses.
H 0: ρ S = 0
H 1: ρ S > 0
762
A B C D
2
3 Parking and Visits
5 z Stat 4.47
z = 4.47; p-value = 0. There is enough evidence to infer that the ratings of parking and the number
of visits is positively related. Thus, there is enough evidence to conclude that the problem of
parking is one reason for the decline in downtown shopping.
H0 : ρ = 0
H1 : ρ > 0
A B
1 Correlation
2
3 Visits and Income
5 t Stat 2.48
6 df 195
t = 2.48; p-value = .0071. There is enough evidence to infer that more affluent people shop more
downtown than poorer people.
A19.7
The histograms indicate that the scores are extremely nonnormal.
Interpersonal skills: Wilcoxon rank sum test

H0: The two population locations are the same
H1: The location of population 1 is to the right of the location of population 2
763
A B C D
2
4 Gp 1 Interpersonal 36392 187
5 Gp 2 Interpersonal 30769 179
6 z Stat 2.05
z = 2.05; p-value = .0200. There is enough evidence to conclude that students whose university
GPA is greater than 3.0 score higher on interpersonal skills than students whose university GPA is
less than 2.
The histograms indicate that the scores are extremely nonnormal.
Stress management: Wilcoxon rank sum test

H1: The location of population 1 is to the right of the location of population 2
A B C D
2
4 Gp 1 Stress Mgmt 36131.5 187
5 Gp 2 Stress Mgmt 31029.5 179
6 z Stat 1.80
z = 1.80; p-value = .0363. There is enough evidence to conclude that students whose university
GPA is greater than 3.0 score higher on stress management than students whose university GPA is
less than 2.
A19.8 All histograms (not shown) are somewhat bell shaped.

One-way analysis of variance
H 0 : µ1 = µ 2 = µ 3
764
Miami-Dade
A B C D E F G
10 ANOVA
12 Between Groups 6,409,467,776 2 3,204,733,888 231.37 1.1E-57 3.03
13 Within Groups 3,476,645,492 251 13,851,177
14
15 Total 9,886,113,268 253
F = 231.37; p-value – 0. There is overwhelming evidence to infer that there are differences
between the three groups of Americans residing in Miami-Dade.
Florida
A B C D E F G
10 ANOVA
12 Between Groups 4,160,159,751 2 2,080,079,875 99.21 7.62E-34 3.03
13 Within Groups 6,331,923,528 302 20,966,634
14
15 Total 10,492,083,279 304
F = 99.21; p-value – 0. There is overwhelming evidence to infer that there are differences between
the three groups of Americans residing in the state of Florida.
United States
A B C D E F G
10 ANOVA
12 Between Groups 5,742,495,149 2 2,871,247,574 110.48 9.27E-41 3.01
13 Within Groups 13,618,559,386 524 25,989,617
14
15 Total 19,361,054,535 526
F = 110.48; p-value – 0. There is overwhelming evidence to infer that there are differences
between the three groups of Americans residing in the United States.
Cars
A B C D
1 t-Estimate: Mean
2
3 Cars
4 Mean 12.20
6 LCL 11.86
7 UCL 12.55
Estimate of the total number of miles driven by cars:
765
LCL = 136 million(11.86 thousand) = 1,613 billion miles
UCL = 136 million(12.55 thousand) = 1,707 billion miles
Buses
A B C D
1 t-Estimate: Mean
2
3 Buses
4 Mean 8.50
6 LCL 8.11
7 UCL 8.88
Estimate of the total number of miles driven by buses:
LCL = 776 thousand (8.11 thousand) = 6,293 million miles
UCL = 776 thousand (8.88thousand) = 6,891 million miles
Vans, pickups, SUV’s

A B C D
1 t-Estimate: Mean
2
3 Vans, pickups, SUVs
4 Mean 11.50
6 LCL 11.09
7 UCL 11.91
Estimate of the total number of miles driven by vans, pickups, SUV’s
LCL = 87 million (11.09 thousand) = 965 billion miles
UCL = 87 million (11.91 thousand) = 1,036 billion miles
A19.10 All three histograms (not shown) are bell shaped.

One-way analysis of variance
H 0 : µ1 = µ 2 = µ 3
A B C D E F G
10 ANOVA
12 Between Groups 1116.46 2 558.23 73.45 3.90E-29 3.01
13 Within Groups 4369.92 575 7.60
14
15 Total 5486.38 577
F = 73.45; p-value = 0. There is overwhelming evidence to infer that there are differences in
distance driven between cars, buses, and vans, pickups, and SUV’s.
766
A B C D
1 t-Estimate: Mean
2
3 Debt
4 Mean 72.15
6 LCL 69.11
7 UCL 75.18
Estimate of total debt:
LCL = 10 million(69.11 thousand) = $691 billion
UCL = 10 million(75.18 thousand) = $751.8 billion
A19.12 Both histograms (not shown) are positively skewed but not sufficiently so to violate the
normality requirement of the t-test of µ1 − µ2.
Equal-variances t-test of µ1 − µ2
H0: (µ1 – µ2) = 0
H1: (µ1 – µ2) < 0
A B C
2
3 5 years ago This year
4 Mean 7.14 7.81
5 Variance 12.87 18.64
9 df 173
10 t Stat -1.11
11 P(T<=t) one-tail 0.1338
13 P(T<=t) two-tail 0.2677
t = − 1.11; p-value = .1338. There is not enough evidence to allow us to infer that investors’
portfolios are becoming more diverse.
A19.13 Both histograms (not shown) are bell shaped.

Equal-variances t-test of µ1 − µ2
H0: (µ1 – µ2) = 5
H1: (µ1 – µ2) > 5
767
A B C
2
3 MBA BBA
4 Mean 96.53 89.08
5 Variance 156.70 106.44
9 df 198
10 t Stat 1.51
11 P(T<=t) one-tail 0.0665
13 P(T<=t) two-tail 0.1330
t = 1.51; p-value = .0665. There is not enough evidence to infer that an MBA is worthwhile.

A B C D E F G H
1 Contingency Table
2
3 Group
4 Side Effect 1 2 3 4 5 TOTAL
5 1 11 10 10 30 106 167
6 2 126 61 54 101 440 782
7 3 111 55 41 78 473 758
8 TOTAL 248 126 105 209 1019 1707
9
10
12 df 8
13 p-value 0.0082
χ 2 = 20.64; p-value = .0082. There is not enough evidence to conclude that there are differences
in side effects between the three groups.
768
A19.15 Multiple regression
A B C D E F
1 SUMMARY OUTPUT
2
4 Multiple R 0.8189
5 R Square 0.6705
8 Observations 100
9
10 ANOVA
12 Regression 6 1667.21 277.87 31.54 2.01E-20
13 Residual 93 819.23 8.81
14 Total 99 2486.44
15
17 Intercept 1.005 1.387 0.725 0.4704
18 Dexterity -0.059 0.150 -0.393 0.6953
19 Detail -0.177 0.159 -1.108 0.2708
20 Teamwork 0.182 0.157 1.161 0.2485
21 Math 0.065 0.162 0.404 0.6874
22 ProbSolve 1.002 0.160 6.268 1.13E-08
23 Tech 1.782 0.159 11.239 5.10E-19
Only problems-solving skills and technical knowledge are linearly related to quality.
A19.16 a t-test of µ
H0 : µ =0
H1 : µ > 0
A B C D
1 t-Test: Mean
2
3 Decrease
4 Mean 24.73
7 df 222
8 t Stat 20.61
9 P(T<=t) one-tail 0
11 P(T<=t) two-tail 0
t = 20.61; p-value = 0. There is overwhelming evidence to infer that there is a decrease in
metabolism when children watch television.
769
b Both histograms (not shown) are roughly bell shaped.
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) > 0
A B C
2
3 Obese Nonobese
4 Mean 30.86 23.35
5 Variance 112.85 358.58
8 df 106
9 t Stat 3.46
10 P(T<=t) one-tail 0.0004
12 P(T<=t) two-tail 0.0008
t = 3.46; p-value = .0004. There is enough evidence to conclude that the decrease in metabolism is
greater among obese children.
A19.17 a z-test of p (success = mathematics major wins)

H 0 : p = .5
H1 : p > .5
A B C D
2
3 Math win
5 Observations 500
7 z Stat 2.59
10 P(Z<=z) two-tail 0.0094
z = 2.59; p-value = .0047. There is enough evidence to infer that mathematics majors win more
frequently than English majors.
bThe matched pairs differences are extremely nonnormal.
770
Wilcoxon signed rank sum test
H 0 : The two population locations are the same.
H1 : The location of population 1 is to the right of the location of population 2.
A B C D
2
3 Difference English - Math
4
5 T+ 67513.5
6 T- 57736.5
8 z Stat 1.512
11 P(Z<=z) two-tail 0.1304
z = 1.512, p-value = .0652. There is not enough evidence to infer that English majors outscore
mathematics majors.
c Part (a) ignores the magnitude of the paired differences.
A19.18a Histograms (not shown) are bell shaped.

H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
t = –3.71; p-value = .0001. There is enough evidence to infer that children who wash their hands
four or more times per day have less sick days due to cold and flu.
b Histograms (not shown) are bell shaped.
771
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Four or more Less
4 Mean 1.76 3.17
8 df 174
9 t Stat -6.52
10 P(T<=t) one-tail 0.0000
12 P(T<=t) two-tail 0.0000
t = –6.52; p-value = 0. There is enough evidence to infer that children who wash their hands four
or more times per day have less sick days due to stomach illness.
H 0: ρ S = 0
H 1: ρ S > 0
A B C D
2
3 EI and Satisfaction
5 z Stat 2.55
z = 2.55; p-value = .0054. There is evidence to conclude that workers who use EI more often are
more satisfied with their employment situation.
A19.20 a Histograms (not shown) of sick days are bell shaped.

772
H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
A B C
2
3 Days flu shot Days placebo
4 Mean 2.82 3.22
8 df 281
9 t Stat -2.69
10 P(T<=t) one-tail 0.0038
12 P(T<=t) two-tail 0.0076
t = –2.69; p-value = .0038. There is overwhelming evidence to indicate that the number of sick
days is less for those who take the flu shots.
b The number of visits is extremely nonnormal.
Wilcoxon rank sum test
A B C D
2
4 Visits flu shot 19152 150
5 Visits placebo 25998 150
6 z Stat -4.56
z = –4.56; p-value = 0. There is overwhelming evidence to indicate that those who take the flu
shots visit their doctors less frequently

a H 0 : ρS = 0
H1 : ρ S ≠ 0
773
A B C D
2
3 Satisfied and Severity
5 z Stat -3.90
z = –3.90; p-value = .0001. There is overwhelming evidence to conclude that satisfaction level is
affected by severity of illness.
b H 0 : ρS = 0
H1 : ρ S < 0
A B C D
2
3 Satisfied and Days
5 z Stat -5.76
z = –5.76; p-value = 0. There is overwhelming evidence to conclude that satisfaction level is
higher for patients who stay for shorter periods of time.

774
A B C D E
1 Contingency Table
2
3 City
4 Outcome 1 2 TOTAL
5 1 503 201 704
6 2 536 215 751
7 3 308 83 391
8 TOTAL 1347 499 1846
9
10
12 df 2
13 p-value 0.0145
χ 2 = 8.47; p-value = .0145. There is enough evidence to conclude that there are differences in the
death rate between the three cities. Recommendation: Spend less time at accident scene and get
patient to the hospital as soon as possible.
A19.23 a z-estimator of p
A B
2 Golfer?
4 Observations 1116
5 LCL 0.1372
6 UCL 0.1800
Estimate of the total number of golfers:
b. z-estimator of p
There were 53 people in the highest income category who did not play golf and 68 who did.
A B C D E
2
4 Sample size 121 0.562 ± 0.088
Estimate of the number of golfers who earn at least $75,000:
775
c Chi-squared test of a contingency table

A B C D E
1 Contingency Table
2
3 Income
4 Golfer? 1 2 TOTAL
5 1 208 6 214
6 2 169 7 176
7 3 148 12 160
8 4 172 30 202
9 5 189 54 243
10 6 53 68 121
11 TOTAL 939 177 1116
12
13
15 df 5
16 p-value 0
χ 2 = 209.4; p-value = 0. There is enough evidence to conclude that income is a determinant in

who plays golf.

776
A B C D E
1 Contingency Table
2
3 Income
4 University? 1 2 TOTAL
5 1 101 29 130
6 2 261 87 348
7 3 40 25 65
8 TOTAL 402 141 543
9
10
12 df 2
13 p-value 0.0417
χ 2 = 6.35; p-value = .0417. There is enough evidence to conclude that family income affects
whether children attend university.
A19.25a Length of stays is extremely nonnormal.

A B C D E
2
4 Sunny stay 22655.5 143
5 Dim stay 39472.5 209
6 z Stat -2.76
z = -2.76; p-value = .0029. There is enough evidence to conclude that the length of stay is smaller
in bright rooms than in dim ones.
b. Amount of pain medication is extremely nonnormal.

777
A B C D E
2
4 Sunny med 21197.5 143
5 Dim med 40930.5 209
6 z Stat -4.31
z = −4.31; p-value = 0. There is enough evidence to conclude that the amount of pain medication
is lower in bright rooms than in dim ones.
H 0: ρ S = 0
H 1: ρ S > 0
z = -1.53. There is no evidence to conclude that younger people believe that that the rich should
pay a bigger share of taxes. In fact, the contrary may be true.
A19.27 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) > 0
1 = Success
778
z = 1.71, p-value = .0438. There is enough evidence to infer that people in 2008 were less likely to
answer the question correctly
A19.28 t-test of ρ
H0 : ρ = 0
H1 : ρ ≠ 0
t = 5.53, p-value = 0. There is enough evidence to conclude that income and job tenure are linearly
related.
H 0: ρ S = 0
H 1: ρ S > 0
779
z = 2.50, p-value = .0061. There is enough evidence to conclude that less educated people are
more likely to believe that when their children are their age their standard of living will be better.
A19.30 H 0 : The locations of all 4 populations are the same.

Histograms reveal that CUREMPYR is not normally distributed. The appropriate technique is
Kruskal-Wallis test.
H = 19.74, p-value = .0002. There is enough evidence to infer that there are differences in job
tenure between the four classes.
A19.31 t-test of ρ
H0 : ρ = 0
H1 : ρ > 0
780
t = 1.03, p-value = .1505. There is not enough evidence to conclude that if you work for a
company for a long time you are more likely to work overtime.
H = 42.09, p-value = 0. There is enough evidence to conclude that there are differences between
the races with respect to the question about their children’s standard of living.
781
χ2 = 9.31, p-value = .0095. There is enough evidence to infer that the three groups of conservatives
differ in their support of capital punishment.
A19.34 t-test of ρ
H0 : ρ = 0
H1 : ρ > 0
z = 4.65. There is no evidence to support the hypothesis that older people are more likely to agree.
In fact, there is evidence to the contrary.
A19.35 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
1 = Success
782
z = -3.74, p-value = .0001. There is enough evidence to infer that people who favor capital
punishment are less likely to favor requiring a police permit to buy a gun.
A19.36 t-test of ρ
H0 : ρ = 0
H1 : ρ > 0
relationship between occupation prestige scores of married couples.
A19.37 Histograms are approximately bell shaped.

H 0 : (µ1 − µ 2 ) = 0
H1 : (µ1 − µ 2 ) < 0
783
t = 8.26, p-value = 4.60E-15 ≈ 0. There is enough evidence to infer that people who own shares in
the company where they work have higher incomes.
A19.38 Equal-variances t-test of µ1 − µ2

H0: (µ1 – µ2) = 0
H1: (µ1 – µ2) < 0
t = -1.13, p-value = .1295. There is not enough evidence to conclude that hearing disabled people
are likely to believe that it is the government’s responsibility to help pay for doctor and hospital
bills.
784
χ2 = 18.77, p-value = .0046. There is enough evidence to conclude that people whose financial
condition worsened in the previous year differ in their political views from those whose financial
condition did not worsen.
A19.40 H 0 : (p1 − p 2 ) = 0
H1 : (p1 − p 2 ) < 0
1 = Success
z = -1.44, p-value = .0747. There is not enough evidence to conclude that victims of a personal
assault are less likely to favor requiring a police permit to buy a gun.
A19.41 Wilcoxon rank sum test

785
z = .635. There is not enough evidence to infer that men believe that they are more likely to lose
their jobs. There is some evidence to the contrary.
H = 30.32, p-value = 0. There is enough evidence to infer that liberals, moderates, and
conservatives differ in their belief that the rich should pay bigger share of taxes.

786
z = .252, p-value = .8014. There is not enough evidence to conclude that men and women differ in
their responses to the question about their standard of living compared to their parents.
A19.44 Simple linear regression
t = 15.06, p-value = 9.61E-47. There is evidence of a linear relationship. For each additional year
of education, income increases on average by $5318.
H 0: ρ S = 0
H 1: ρ S > 0
787
z = 13.48, p-value = 0. There is enough evidence to infer that those who believe the economy has
gotten worse also believe that inflation has gotten worse.
H = 32.25, p-value = 0. There is enough evidence to infer that liberals, moderates, and
conservatives differ in their perception about the level of unemployment in the previous year.
A19.47 Sign test

H0: The location of the two populations are the same
H1: The location of population 1 is different from the location of population2
788
z = 1.62, p-value = .1046. There is not enough evidence to conclude that that attention paid to
news about national politics in a printed newspaper differs from attention paid to news about
national politics on radio.
H = 14.67, p-value = .0007. There is enough evidence to infer that liberals, moderates, and
conservatives differ in their perception about the level of inflation in the previous year.
H 0: ρ S = 0
H 1: ρ S < 0
789
z = -3.37, p-value = .0004. There is enough evidence to infer that more educated people pay more
attention to national network news about the campaign for president.

z = -.266, p-value = .7904. There is not enough evidence to conclude that Democrats and
Republicans differ in their interest in information about what’s going on in government and
politics.

790
z = .533, p-value = .5938. There is not enough evidence to infer that men and women differ in the
attention they pay to national news about the campaign for president.
791
792
Chapter 20
20.1 Time series Moving average

48
41 (48+41+37)/3 = 42.00
37 (41+37+32)/3 = 36.67
32 (37+32+36)/3 = 35.00
36 (32+36+31)/3 = 33.00
31 (36+31+43)/3 = 36.67
43 (31+43+52)/3 = 42.00
52 (43+52+60)/3 = 51.67
60 (52+60+48)/3 = 53.33
48 (60+48+41)/3 = 49.67
41 (48+41+30)/3 = 39.67
30

48
41
37 (48 +41+37+32+36)/5 = 38.8
32 (41+37+32+36+31)/5 = 35.4
36 (37+32+36+31+43)/5 = 35.8
31 (32+36+31+43+52)/5 = 38.8
43 (36+31+43+52+60)/5 = 44.4
52 (31+43+52+60+48)/5 = 46.8
60 (43+52+60+48+41)/5 = 48.8
48 (52+60+48+41+30)/5 = 46.2
41
30
20.3
70
60 Series1
50
Series2
40
30
Series3
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12
793
16
22 (16+22+19)/3 = 19.00
19 (22+19+24)/3 = 21.67
24 (19+24+30)/3 = 24.33
30 (24+30+26)/3 = 26.67
26 (30+26+24)/3 = 26.67
24 (26+24+29)/3 = 26.33
29 (24+29+21)/3 = 24.67
21 (29+21+23)/3 = 24.33
23 (21+23+19)/3 = 21.00
19 (23+19+15)/3 = 19.00
15

16
22
19 (16+22+19+24+30)/5 = 22.2
24 (22+19+24+30+26)/5 = 24.2
30 (19+24+30+26+24)/5 = 24.6
26 (24+30+26+24+29)/5 = 26.6
24 (30+26+24+29+21)/5 = 26.0
29 (26+24+29+21+23)/5 = 24.6
21 (24+29+21+23+19)/5 = 23.2
23 (29+21+23+19+15)/5 = 21.4
19
15
20.6
35
30 Series1
25
Series2
20
15
Series3
10
0
1 2 3 4 5 6 7 8 9 10 11 12
794
20.7 Time series Exponentially smoothed time series
12 12
18 .1(18) +. 9(12) = 12.60
16 .1(16) +. 9(12.60) = 12.94
24 .1(24) +. 9(12.94) = 14.05
17 .1(17) +. 9(14.05) = 14.34
16 .1(16) +. 9(14.34) = 14.51
25 .1(25) +. 9(14.51) = 15.56
21 .1(21) +. 9(15.56) = 16.10
23 .1(23) + .9(16.10) = 16.79
14 .1(14) + .9(16.79) = 16.51

12 12.00
18 .8(18) +. 2(12) = 16.80
16 .8(16) +. 2(16.80) = 16.16
24 .8(24) +. 2(16.16) = 22.43
17 .8(17) +. 2(22.43) = 18.09
16 .8(16) +. 2(18.09) = 16.42
25 .8(25) +. 2(16.42) = 23.28
21 .8(21) +. 2(23.28) = 21.46
23 .8(23) + .2(21.46) = 22.69
14 .8(14) + .2(22.69) = 15.74
20.9
30
25 Series1
20
Series2
15
Series3
10
0
1 2 3 4 5 6 7 8 9 10
There appears to be a gradual upward trend.
795
38 38.00
43 .1(43) +. 9(38) = 38.50
42 .1(42) +. 9(38.50) = 38.85
45 .1(45) +. 9(38.85) = 39.47
46 .1(46) +. 9(39.47) = 40.12
48 .1(48) +. 9(40.12) = 40.91
50 .1(50) +. 9(40.91) = 41.82
49 .1(49) +. 9(41.82) = 42.53
46 .1(46) + .9(42.53) = 42.88
45 .1(45) + .9(42.88) = 43.09
There appears to be a gradual upward trend.

38 38
43 .8(43) +. 2(38) = 42.00
42 .8(42) +. 2(42.00) = 42.00
45 .8(45) +. 2(42.00) = 44.40
46 .8(46) +. 2(44.40) = 45.68
48 .8(48) +. 2(45.68) = 47.54
50 .8(50) +. 2(47.54) = 49.51
49 .8(49) +. 2(49.51) = 49.10
46 .8(46) + .2(49.10) = 46.62
45 .8(45) + .2(46.62) = 45.32
20.12
55
Series1
50
Series2
45
40 Series3
35
30
1 2 3 4 5 6 7 8 9 10
There is a trend component.
796
20.13 & 20.14
Sales 3-Day moving average 5-Day moving average
43
45 (43+45+22)/3 = 36.67
22 (45+22+25)/3 = 30.67 (43+45+22+25+31)/5 = 33.20
25 (22+25+31)/3 = 26.00 (45+22+25+31+51)/5 = 34.80
31 (25+31+51)/3 = 35.67 (22+25+31+51+41)/5 = 34.00
51 (31+51+41)/3 = 41.00 (25+31+51+41+37)/5 = 37.00
41 (51+41+37)/3 = 43.00 (31+51+41+37+22)/5 = 36.40
37 (41+37+22)/3 = 33.33 (51+41+37+22+25)/5 = 35.20
22 (37+22+25)/3 = 28.00 (41+37+22+25+40)/5 = 33.00
25 (22+25+40)/3 = 29.00 (37+22+25+40+57)/5 = 36.20
40 (25+40+57)/3 = 40.67 (22+25+40+57+30)/5 = 34.80
57 (40+57+30)/3 = 42.33 (25+40+57+30+33)/5 = 37.00
30 (57+30+33)/3 = 40.00 (40+57+30+33+37)/5 = 39.40
33 (30+33+37)/3 = 33.33 (57+30+33+37+64)/5 = 44.20
37 (33+37+64)/3 = 44.67 (30+33+37+64+58)/5 = 44.40
64 (37+64+58)/3 = 53.00 (33+37+64+58+33)/5 = 45.00
58 (64+58+33)/3 = 51.67 (37+64+58+33+38)/5 = 46.00
33 (58+33+38)/3 = 43.00 (64+58+33+38+25)/5 = 43.60
38 (33+38+25)/3 = 32.00
25
70
60 Series1
50
Series2
40
30
Series3
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
c There appears to be a seasonal (weekly) pattern.
797
20.15a Sales 4-quarter moving average Centered moving average
18
22
(18+22+27+31)/4 = 24.50
27 (24.50+28.25)/2 = 26.375
(22+27+31+33)/4 = 28.25
31 (28.25+27.75)/2 = 28.000
(27+31+33+20)/4 = 27.75
33 (27.75+30.50)/2 = 29.125
(31+33+20+38)/4 = 30.50
20 (30.50+29.25)/2 = 29.875
(33+20+38+26)/4 = 29.25
38 (29.25+27.25)/2 = 28.250
(20+38+26+25)/4 = 27.25
26 (27.25+31.25)/2 = 29.250
(38+26+25+36)/4 = 31.25
25 (31.25+32.75)/2 = 32.000
(26+25+36+44)/4 = 32.75
36 (32.75+33.50)/2 = 33.125
(25+36+44+29)/4 = 33.50
44 (33.50+37.50)/2 = 35.500
(36+44+29+41)/4 = 37.50
29 (37.50+36.75)/2 = 37.125
(44+29+41+33)/4 = 36.75
41 (36.75+38.75)/2 = 37.750
(29+41+33+52)/4 = 38.75
33 (38.75+42.75)/2 = 40.750
(41+33+52+25)/4 = 42.75
52
45
60
50
Series1
40
30 Series2
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
c There appears to be a gradual trend of increasing sales.
798
20.16 & 20.17
Sales Exponentially smoothed w = .4 Exponentially smoothed w = .8
18 18.00 18.00
22 .4(22)+.6(18) = 19.60 .8(22)+.2(18) = 21.20
27 .4(27)+.6(19.6) = 22.56 .8(27)+.2(21.2) = 25.84
31 .4(31)+.6(22.56) = 25.94 .8(31)+.2(25.84) = 29.97
33 .4(33)+.6(25.94) = 28.76 .8(33)+.2(29.97) = 32.39
20 .4(20)+.6(28.76) = 25.26 .8(20)+.2(32.39) = 22.48
38 .4(38)+.6(25.26) = 30.35 .8(38)+.2(22.48) = 34.90
26 .4(26)+.6(30.35) = 28.61 .8(26)+.2(34.90) = 27.78
25 .4(25)+.6(28.61) = 27.17 .8(25)+.2(27.78) = 25.56
36 .4(36)+.6(27.17) = 30.70 .8(36)+.2(25.56) = 33.91
44 .4(44)+.6(30.70) = 36.02 .8(44)+.2(33.91) = 41.98
29 .4(29)+.6(36.02) = 33.21 .8(29)+.2(41.98) = 31.60
41 .4(41)+.6(33.21) = 36.33 .8(41)+.2(31.60) = 39.12
33 .4(33)+.6(36.33) = 35.00 .8(33)+.2(39.12) = 34.22
52 .4(52)+.6(35.00) = 41.80 .8(52)+.2(34.22) = 48.44
45 .4(45)+.6(41.80) = 43.08 .8(45)+.2(48.44) = 45.69
20.16 b
60
50 Series1
40
30 Series2
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
799
20.17 b
60
50
Series1
40
30 Series2
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
20.18
Line chart
25
20
Time series
15
10
0
1 2 3 4 5 6 7 8
Period
The quadratic model would appear to be the best model.
800
20.19
Line chart
60
50
Time series
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10
Period
The linear trend model appears to be best.
20.20
ŷ = −4.96 + 2.38t (R 2 = .81)

ŷ = 3.14 − 2.48t + .54t 2 (R 2 = .98)
The quadratic trend line fits better.
20.21
ŷ = 63.87 − 3.94t (R 2 = .94)

ŷ = 57.2 − .61t − .30t 2 (R 2 = .98)
The quadratic trend line fits slightly better.
801
20.22
Week Day Period t y ŷ y / yˆ

1 1 1 12 17.2 0.699
2 2 18 17.5 1.027
3 3 16 17.9 0.894
4 4 25 18.3 1.369
5 5 31 18.6 1.664
2 1 6 11 19.0 0.579
2 7 17 19.4 0.878
3 8 19 19.7 0.963
4 9 24 20.1 1.194
5 10 27 20.5 1.320
3 1 11 14 20.8 0.672
2 12 16 21.2 0.755
3 13 16 21.6 0.742
4 14 28 21.9 1.277
5 15 25 22.3 1.122
4 1 16 17 22.7 0.750
2 17 21 23.0 0.912
3 18 20 23.4 0.855
4 19 24 23.8 1.010
5 20 32 24.1 1.327
Day
Week Monday Tuesday Wednesday Thursday Friday Total
1 .699 1.027 .894 1.369 1.664
2 .579 .878 .963 1.194 1.320
3 .672 .755 .742 1.277 1.122
4 .750 .912 .855 1.010 1.327
Average .675 .893 .864 1.213 1.358 5.003
Seasonal
Index .675 .892 .864 1.212 1.357 5.000
802
20.23
Year Quarter y ŷ y / yˆ
1 1 55 46.6 1.179
2 44 45.6 0.965
3 46 44.5 1.033
4 39 43.5 0.897
2 1 41 42.4 0.967
2 38 41.3 0.919
3 37 40.3 0.919
4 30 39.2 0.765
3 1 43 38.2 1.127
2 39 37.1 1.051
3 39 36.0 1.082
4 35 35.0 1.001
4 1 36 33.9 1.061
2 32 32.9 0.974
3 30 31.8 0.943
4 25 30.7 0.813
5 1 50 29.7 1.685
2 25 28.6 0.874
3 24 27.6 0.871
4 22 26.5 0.830
Quarter
Year 1 2 3 4 Total
1 1.179 0.965 1.033 0.897
2 0.967 0.919 0.919 0.765
3 1.127 1.051 1.082 1.001
4 1.061 0.974 0.943 0.813
5 1.685 0.874 0.871 0.830
Average 1.204 0.957 0.970 0.861 3.991
Seasonal
Index 1.207 0.959 0.972 0.863 4.000
803
20.24
Year Quarter Period t y ŷ y / yˆ
2001 1 1 52 62.9 0.827
2 2 67 64.1 1.045
3 3 85 65.3 1.302
4 4 54 66.5 0.812
2002 1 5 57 67.7 0.842
2 6 75 68.8 1.090
3 7 90 70.0 1.286
4 8 61 71.2 0.857
2003 1 9 60 72.4 0.829
2 10 77 73.6 1.046
3 11 94 74.7 1.258
4 12 63 75.9 0.830
2004 1 13 66 77.1 0.856
2 14 82 78.3 1.047
3 15 98 79.5 1.233
4 16 67 80.6 0.831
Quarter
Year 1 2 3 4 Total
2001 .827 1.045 1.302 .812
2002 .842 1.090 1.286 .857
2003 .829 1.046 1.258 .830
2004 .856 1.047 1.233 .831
Average .838 1.057 1.270 .833 3.998
Seasonal
Index .839 1.058 1.270 .833 4.000
20.25a
b ŷ = −491,595 + 253.78 Year
804
20.26a
b ŷ = 49,661 – 25.039Year
20.27
Scatter Diagram
400
y = 7.416x + 143.0
350
300
Subscribers
250
200
150
100
50
0
0 5 10 15 20 25 30
Quarter
ŷ = 143 + 7.42 t
805
Year Quarter Period t y ŷ =143+7.42t y / yˆ
1 1 1 184 150.42 1.223
2 2 173 157.84 1.096
3 3 160 165.26 0.968
4 4 189 172.68 1.095
2 1 5 191 180.1 1.061
2 6 185 187.52 0.987
3 7 184 194.94 0.944
4 8 200 202.36 0.988
3 1 9 205 209.78 0.977
2 10 192 217.2 0.884
3 11 200 224.62 0.890
4 12 229 232.04 0.987
4 1 13 236 239.46 0.986
2 14 219 246.88 0.887
3 15 211 254.3 0.830
4 16 272 261.72 1.039
5 1 17 280 269.14 1.040
2 18 261 276.56 0.944
3 19 275 283.98 0.968
4 20 322 291.4 1.105
6 1 21 331 298.82 1.108
2 22 301 306.24 0.983
3 23 306 313.66 0.976
4 24 351 321.08 1.093
Quarter
Year 1 2 3 4 Total
1 1.223 1.096 .968 1.095
2 1.060 .987 .944 .988
3 .977 .884 .890 .987
4 .986 .887 .830 1.039
5 1.040 .944 .968 1.105
6 1.108 .983 .976 1.093
Average 1.066 .964 .929 1.051 4.010
Seasonal
Index 1.063 .962 .927 1.048 4.000
20.28 Regression line: ŷ = 145 + 1.66 t
806
Week Day Period t y ŷ =145+1.66t y / yˆ
1 1 1 240 146.66 1.636
2 2 85 148.32 0.573
3 3 93 149.98 0.620
4 4 106 151.64 0.699
5 5 125 153.3 0.815
6 6 188 154.96 1.213
7 7 314 156.62 2.005
2 1 8 221 158.28 1.396
2 9 80 159.94 0.500
3 10 75 161.6 0.464
4 11 121 163.26 0.741
5 12 110 164.92 0.667
6 13 202 166.58 1.213
7 14 386 168.24 2.294
3 1 15 235 169.9 1.383
2 16 86 171.56 0.501
3 17 74 173.22 0.427
4 18 100 174.88 0.572
5 19 117 176.54 0.663
6 20 205 178.2 1.150
7 21 402 179.86 2.235
4 1 22 219 181.52 1.206
2 23 91 183.18 0.497
3 24 102 184.84 0.552
4 25 89 186.5 0.477
5 26 105 188.16 0.558
6 27 192 189.82 1.011
7 28 377 191.48 1.969
Day
Week Sunday Monday Tuesday Wednesday Thursday Friday Saturday Total
1 1.636 .573 .620 .699 .815 1.213 2.005
2 1.396 .500 .464 .741 .667 1.213 2.294
3 1.383 .501 .427 .572 .663 1.150 2.235
4 1.206 .497 .552 .477 .558 1.011 1.969
Average 1.405 .518 .516 .657 .676 1.187 2.126 7.085
Seasonal
Index 1.404 .517 .515 .621 .675 1.145 2.123 7.000
20.29 Regression line: ŷ = 90.4 + 2.02 t
807
Year Quarter t y ŷ =90.4 + 2.02t y / yˆ
1 1 1 106 92.42 1.147
2 2 92 94.44 0.974
3 3 65 96.46 0.674
4 4 121 98.48 1.229
2 1 5 115 100.5 1.144
2 6 100 102.52 0.975
3 7 73 104.54 0.698
4 8 135 106.56 1.267
3 1 9 114 108.58 1.050
2 10 105 110.6 0.949
3 11 79 112.62 0.701
4 12 140 114.64 1.221
4 1 13 121 116.66 1.037
2 14 111 118.68 0.935
3 15 82 120.7 0.679
4 16 163 122.72 1.328
Quarter
Year 1 2 3 4 Total
1 1.147 .974 .674 1.229
2 1.144 .975 .698 1.267
3 1.050 .949 .701 1.221
4 1.037 .935 .679 1.328
Average 1.095 .959 .688 1.261 4.003
Seasonal
Index 1.094 .958 .688 1.260 4.000
20.30
166 − 173 + 179 − 186 + 195 − 192 + 214 − 211 + 220 − 223
MAD =
5
7 + 7 + 3 + 3 + 3 23
= = = 4.60
5 5
= (166 − 173) 2 + (179 − 186) 2 + (195 − 192) 2 + (214 − 211) 2 + (220 − 223) 2
= 49 + 49 + 9 + 9 + 9 = 125
20.31 Model 1:
6.0 − 7.5 + 6.6 − 6.3 + 7.3 − 5.4 + 9.4 − 8.2
MAD =
4
1.5 + .3 + 1.9 + 1.2
=
4
4 .9
= = 1.225
4
808
Model 2:
6.0 − 6.3 + 6.6 − 6.7 + 7.3 − 7.1 + 9.4 − 7.5
MAD =
4
.3 + .1 + .2 + 1.9 2.5
= = = .625
4 4
Model 2 is more accurate.
Model 1: SSE = (6.0 - 7.5)2 + (6.6 - 6.3)2 + (7.3 - 5.4)2 + (9.4 - 8.2)2
= 2.25 + .09 + 3.61 + 1.44 = 7.39
Model 2: SSE = (6.0 - 6.3)2 + (6.6 - 6.7)2 + 7.3 - 7.1)2 + (9.4 - 7.5)2
= .09 + .01 + .04 + 3.61 = 3.75
Model 2 is more accurate.
20.32
57 − 63 + 60 − 72 + 70 − 86 + 75 − 71 + 70 − 60
MAD =
5
6 + 12 + 16 + 4 + 10 48
= = = 9.6
5 5
SSE = (57 - 63)2 + (60 - 72)2 + (70 - 86)2 + (75 - 71)2 +(70-60)2
= 36 + 144 + 256 + 16 + 100 = 552
20.33 Technique 1:
19 − 21 + 24 − 27 + 28 − 29 + 32 − 31 + 38 − 35
MAD =
5
2 + 3 + 1 + 1 + 3 10
= = = 2.0
5 5
SSE = (19 - 22)2 + (24 - 27)2 + (32 - 31)2 + (38 - 35)2
= 4 + 9 + 1 + 1 + 9 = 24
Technique 2:
19 − 22 + 24 − 24 + 28 − 26 + 32 − 28 + 38 − 30
MAD =
5
3 + 0 + 2 + 4 + 8 15
= = = 3.0
5 5
SSE = (19 - 22)2 + (24 - 24)2 + (28 - 26)2 + (32 - 28)2 + (28 - 30)2
= 9 + 0 + 4 + 16 + 64 = 93
809
Technique 3:
19 − 17 + 24 − 20 + 28 − 25 + 32 − 31 + 38 − 39
MAD =
5
2 + 4 + 3 + 1 + 1 11
= = = 2.2
5 5
SSE = (19 - 17)2 + (24 - 20)2 + (28 - 25)2 + (32 - 31)2 + (38 - 39)2
= 4 + 16 + 9 + 1+ 1 = 31
By both measures, technique 1 is the most accurate.
20.34
Quarter t yˆ = 150 + 3t SI Forecast
1 41 273 .7 191.1
2 42 276 1.2 331.2
3 43 279 1.5 418.5
4 44 282 .6 169.2
20.35
Day t yˆ = 120 + 2.3t SI Forecast
Sunday 29 186.7 1.5 280.1
Monday 30 189.0 .4 75.6
Tuesday 31 191.3 .5 95.7
Wednesday 32 193.6 .6 116.2
Thursday 33 195.9 .7 137.1
Friday 34 198.2 1.4 277.5
Saturday 35 200.5 1.9 381.0
20.36 yˆ t = 625 −1.3 yt −1 = 625 −1.3(65) = 540.5
20.37 yˆ t = 155 + 21 y = 155 + 21(11) = 386

t −1
20.38 F17 = F18 = F19 = F20 = S16 = 43.08
20.39
Day t yˆ = 16.8 + .366t SI Forecast
1 21 24.49 .675 16.53
2 22 24.85 .892 22.17
3 23 25.22 .864 21.79
4 24 25.58 1.212 31.01
5 25 25.95 1.357 35.21
810
20.40
Quarter t yˆ = 47.7 − 1.06t SI Forecast
1 21 25.44 1.207 30.71
2 22 24.38 .959 23.38
3 23 23.32 .972 22.67
4 24 22.26 .863 19.21
20.41
Quarter t yˆ = 61.75 + 1.18t SI Forecast
1 17 81.81 0.839 68.64
2 18 82.99 1.058 87.80
3 19 84.17 1.270 106.90
4 20 85.35 0.833 71.10
1 21 86.53 0.839 72.60
2 22 87.71 1.058 92.80
3 23 88.89 1.270 112.89
4 24 90.07 0.833 75.03
20.42a ŷ 2007 = 1,775.2 + .9054y 2006 = 1,775.2 + .9054(17,672) = 17,775
b F2007 = S2006 = 17,146.
20.43 a ŷ 2007 = −4.2245 + 1.12032006 = −4.2245 + 1.1203(−817.3) = −919.8
F2007 = S2006 = −719.5
20.44
Quarter t yˆ = 143 + 7.42t SI Forecast
1 25 328.50 1.063 349.20
2 26 335.92 .962 323.16
3 27 343.34 .927 318.28
4 28 350.76 1.048 367.60
20.45
Day t yˆ = 145 + 1.66t SI Forecast
1 29 193.14 1.404 271.17
2 30 194.80 0.517 100.71
3 31 196.46 0.515 101.18
4 32 198.12 0.621 123.03
5 33 199.78 0.675 134.85
6 34 201.44 1.145 230.65
7 35 203.10 2.123 431.18
811
20.46
Day t yˆ = 90.4 + 2.02t SI Forecast
1 17 124.74 1.094 136.47
2 18 126.76 0.958 121.44
3 19 128.78 0.688 88.60
4 20 130.80 1.260 164.81
20.47
60
50
40
Revenues
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Quarter
20.48 There is a small upward trend and seasonality.
20.49 ŷ = 20.21 + .732t
20.50
A B
1 Seasonal Indexes
2
3 Season Index
4 1 0.646
5 2 1.045
6 3 1.405
7 4 0.904
20.51
Quarter t yˆ = 20.21 + .732t SI Forecast
1 21 35.58 0.646 23.00
2 22 36.31 1.045 37.95
3 23 37.05 1.405 52.05
4 24 37.78 0.904 34.14
812
20.52a
b.
Period Month ŷ = 16.34-.1008t Seasonal Index Forecasts Actual
61 January 10.19 .7192 7.33 2.2

62 February 10.09 .6729 6.79 3.6
63 March 9.99 .8936 8.93 5.3
64 April 9.89 1.0500 10.38 4.5
65 May 9.79 1.2148 11.89 5.7
66 June 9.69 1.3998 13.56 8.1
67 July 9.59 1.1908 11.42 6.1
68 August 9.49 1.1651 11.05 6.5
69 September 9.38 1.0246 9.62 5.8
70 October 9.28 1.0288 9.55 4.9
71 November 9.18 .8535 7.84 4.8
72 December 9.08 .7870 7.15 4.4
c. MAD = 4.47
SSE = 254.32
813
814
Chapter 21
21.1Chance variation is caused by a number of randomly occurring events that are part of the
production process and that in general cannot be eliminated without changing the process.
21.3 Special variation is caused by specific events or factors that are frequently temporary and that
can usually be identified and eliminated
21.4a Chance variation represents the variation in student achievement caused by differences in
preparation, motivation, and ability.
b Special variation represents variation due to specific event or factors that can be corrected.
21.5 α = P(| z |> 2.5) = 2(1 − .9938) = .0124
1
21.6 ARL = = 81
.0124
21.7 α = P(| z |> 2.0) = 2(1 − .9772) = .0456
1
21.8 ARL = = 22
.0456
1
21.9 ARL = = 385
.0026
Number of units = Production × ARL = 100(385) = 38,500
21.10a From Beta-mean spreadsheet, β = .6603
b Probability = .6603 8 = .0361
1 1
21.11 P = 1 − β = 1 − .6603 = .3397; ARL = = = 2.94
P .3397
21.12 Number of units = Production × ARL = 50(385) = 19,250
b Probability = .8133 8 = .1914
815
1 1
21.14 P = 1 − β = 1 − .8133 = .1867; ARL = = = 5.36
P .1867
21.15 Sampling 3 units per hour means that on average we will produce 38,500 units before
erroneously concluding that the process is out of control when it isn’t. Sampling 2 units per half
hour reduces this figure by 50%. Sampling 4 units per hour means that when the process goes out
of control, the probability of not detecting a shift of 1.5 standard deviations is .6603 and we will
produce on average 2.94 × 100 = 294 units until the chart indicates a problem. Sampling 2 units
per half hour increases the probability of not detecting the shift to .8133 and decreases the average
number of units produced when the process is out of control to 50 × 5.36 = 268.
21.16 Number of units = Production × ARL = 2000(385) = 770,000
b Probability = .7388 4 = .2979
1 1
21.18 P = 1 − β = 1 − .7388 = .2612; ARL = = = 3.83
P .2612
21.19 Number of units = Production × ARL = 4000(385) = 1,540,000
b Probability = .3659 4 = .0179
1 1
21.21 P = 1 − β = 1 − .3659 = .6341; ARL = = = 1.58
P .6341
21.22 Sampling 10 units per half hour means that on average we will produce 770,000 units before
erroneously concluding that the process is out of control when it isn’t. Sampling 20 units per hour
doubles this figure. Sampling 10 units per half hour means that when the process goes out of
control, the probability of not detecting a shift of .75 standard deviations is .7388 and we will
produce on average 3.83 × 2000 = 7660 units until the chart indicates a problem. Sampling 20
units per hour decreases the probability of not detecting the shift to .3659 and decreases the
average number of units produced when the process is out of control to 4000 × 1.58 = 6320.
21.23 Centerline = x = 453.6
816
3S ⎛ 12.5 ⎞
Lower control limit = x − = 453.6 − 3 ⎜⎜ ⎟⎟ = 434.85
n ⎝ 4 ⎠
3S ⎛ 12.5 ⎞
Upper control limit = x + = 453.6 + 3 ⎜⎜ ⎟⎟ = 472.35
n ⎝ 4 ⎠
21.24 Centerline = x = 181.1
3S ⎛ 11.0 ⎞
Lower control limit = x − = 181.1 − 3 ⎜⎜ ⎟ = 170.1
⎟
n ⎝ 9 ⎠
3S ⎛ 11.0 ⎞
Upper control limit = x + = 181.1 + 3 ⎜⎜ ⎟ = 192.1
⎟
n ⎝ 9 ⎠
Zone boundaries: 170.10, 173.77, 177.44, 181.10, 184.77, 188.43, 192.10
21.25 a Centerline = x = 13.3
3S ⎛ 3.8 ⎞
Lower control limit = x − = 13.3 − 3 ⎜⎜ ⎟⎟ = 7.6
n ⎝ 4 ⎠
3S ⎛ 3.8 ⎞
Upper control limit = x + = 13.3 + 3 ⎜⎜ ⎟⎟ = 19.0
n ⎝ 4 ⎠
c The process is out of control at points 8, 9, 21, 22, and 25.
21.26a S Chart
A B C D
1 Statistical Process Control
2
3 Data
4 Upper control limit 10.0885
5 Centerline 4.452
6 Lower control limit 0
817
12
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
x Chart
A B C D
2
3 Data
5 Centerline 12.7386
6 Lower control limit 5.5103
7 Pattern Test #2 Failed at Points: 29
8 Pattern Test #6 Failed at Points: 29, 30
818
25
20
15
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
c The process is out of control at samples 29 and 30.

d A level shift occurred.
21.27a S Chart
A B C
2
3 2by4 lumber
5 Centerline 0.0721
819
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
A B C D E F G
2
3 2by4 lumber
7 Pattern Test #1 Failed at Points: 1, 4, 5, 28, 35, 36, 37, 39, 40
8 Pattern Test #5 Failed at Points: 3, 4, 5, 6, 9, 10, 35, 36, 37, 38, 39, 40
9 Pattern Test #6 Failed at Points: 5, 6, 7, 8, 9, 35, 36, 37, 38, 39, 40
820
96.2
96.1
96
95.9
95.8
95.7
95.6
95.5
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
b The process is out of control at the first sample.

c Clean out the sawdust more frequently.
21.28
A B C
2
3 AEU
5 Centerline 0.0015
821
0.0035
0.003
0.0025
0.002
0.0015
0.001
0.0005
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
A B C
2
3 AEU
5 Centerline 0.4387
0.442
0.441
0.44
0.439
0.438
0.437
0.436
0.435
0.434
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
The process is under control.
822
S Upper control limit − Centerline .4408 − .4387
21.29 = = = .0007
n 3 3
S
= .0007; S = .0016
5
x − LSL .4387 − .4335
CPL = = = 1.08
3S 3(.0016)
USL − x 4435 − .4387

CPU = = = 1.00
3S 3(.0016)
C pk = Min(CPL, CPU) = 1.00
21.30
A B C
2
3 Volume
5 Centerline 0.9093
1.8
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
A B C
2
3 Volume
5 Centerline 0.092
6 Lower control limit -1.2031
823
2
1.5
0.5
-‐0.5
-‐1
-‐1.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
S Upper control limit − Centerline 1001.3871 − 1000.092

21.31 = = = .4317
n 3 3
S
= .4317; S = .9653
5
x − LSL 1000.092 − 995
CPL = = = 1.76
3S 3(.9653)
USL − x 1005 − 1000.092

CPU = = = 1.69
3S 3(.9653)
Cpk = Min(CPL, CPU) = 1.69
21.32
A B C
2
3 Headrest
5 Centerline 0.9078
824
2.5
1.5
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
A B C D
2
3 Headrest
242
241
240
239
238
237
236
235
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
825
a The process is out of control.
b The process is out of control at sample 19.
c The width became too small.
21.33
A B C D
2
3 Nuts
5 Centerline 2.0363
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
The process is out of control at sample 21. It is not necessary to draw the x chart.
21.34
A B C D
2
3 Seats
5 Centerline 5.3523
826
14
12
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
The process is out of control at sample 23. It is not necessary to draw the x chart.
21.35
A B C
2
3 Headers
5 Centerline 0.002
827
0.005
0.004
0.003
0.002
0.001
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
A B C
2
3 Headers
5 Centerline 4.9841
4.988
4.986
4.984
4.982
4.98
4.978
4.976
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
828
21.36 = = = .00107
n 3 3
S
= .00107; S = .00214
4
x − LSL 4.9841 − 4.978
CPL = = = .95
3S 3(.00214)
USL − x 4.990 − 4.9841

CPU = = = .92
3S 3(.00214)
Cpk = Min(CPL, CPU) = .92
21.37
A B C
2
3 Bolts
5 Centerline 0.5021
1.4
1.2
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
829
A B C D E
2
3 Bolts
5 Centerline 11.81
8 Pattern Test #6 Failed at Points: 23, 24, 25
13
12.5
12
11.5
11
10.5
10
9.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
The process went out of control at sample 23.
21.38
A B C
2
3 Bottles
830
80
70
60
50
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
A B C D
2
3 Bottles
850
830
810
790
770
750
730
710
690
670
650
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
831
21.39a
A B C
2
3 Calls
50
45
40
35
30
25
20
15
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
A B C D
2
3 Calls
832
140
120
100
80
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
b The process went out of control at sample 25
21.40
A B C
2
3 Pipes
5 Centerline 0.0372
833
0.12
0.1
0.08
0.06
0.04
0.02
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
A B C
2
3 Pipes
5 Centerline 2.9892
3.1
3.05
2.95
2.9
2.85
2.8
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
834
21.41 = = = .0241
n 3 3
S
= .0241; S = .0417
3
x − LSL 2.9892 − 2.9
CPL = = = .71
3S 3(.0417)
USL − x 3.1 − 2.9892

CPU = = = .89
3S 3(.0417)

21.42 = = = 2.54
n 3 3
S
= 2.54; S = 5.68
5
x − LSL 1496.952 − 1486
CPL = = = .64
3S 3(5.68)
USL − x 1506 − 1496.952

CPU = = = .53
3S 3(5.68)
The value of the index is low because the statistics used to calculate the control limits and
centerline were taken when the process was out of control.
21.43 Centerline = p = .035
p(1 − p) (.035)(1 − .035)

Lower control limit = p − 3 = .035 − 3 = .0174
n 1000
p(1 − p) (.035)(1 − .035)

Upper control limit = p + 3 = .035 + 3 = .0524
n 1000
21.44 Centerline = p = .0324
p(1 − p) (.0324)(1 − .0324)

Lower control limit = p − 3 = .0324 − 3 = −.00516 (= 0)
n 200
p(1 − p) (.0324)(1 − .0324)

n 200
835
A B C D
2
3 Copiers
5 Centerline 0.0324
0.014
0.012
0.01
0.008
0.006
0.004
0.002
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
The process is out of control at sample 25.
21.45 Centerline = p = .00352
p(1 − p) (.00352)(1 − .00352)

Lower control limit = p − 3 = .00352 − 3 = −.00443 (=0)
n 500
p(1 − p) (.00352)(1 − .00352)

n 500
A B C
2
3 PCBs
5 Centerline 0.0035
836
0.012
0.01
0.008
0.006
0.004
0.002
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
21.46 Centerline = p = . 0383
p(1 − p) (.0383)(1 − .0383)

Lower control limit = p − 3 = .0383 − 3 = −.0193 (= 0)
n 100
p(1 − p) (.0383)(1 − .0383)

n 100
A B C D
2
3 Telephones
5 Centerline 0.0383
8
837
0.12
0.1
0.08
0.06
0.04
0.02
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
The process is out of control at samples 25 and 30.
21.47 Centerline = p = .0169
p(1 − p) (.0169)(1 − .0169)

Lower control limit = p − 3 = .0169 − 3 = .0047
n 1000
p(1 − p) (.0169)(1 − .0169)

n 1000
A B C D
2
3 Pages
5 Centerline 0.0169
838
0.035
0.03
0.025
0.02
0.015
0.01
0.005
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
The process is out of control at samples 37-40.
21.48
A B C D E
2
3 Batteries
5 Centerline 0.0257
7 Pattern Test #1 Failed at Points: 28, 29, 30
839
0.06
0.05
0.04
0.03
0.02
0.01
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
21.49
A B C D
2
3 Courier
5 Centerline 0.0044
840
0.01
0.008
0.006
0.004
0.002
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
21.50
A B C D
2
3 Scanners
5 Centerline 0.0126
841
0.035
0.03
0.025
0.02
0.015
0.01
0.005
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
842
Chapter 22
22.1 a1 a2
s1 0 29
s2 0 5
s3 14 0
s4 36 0
22.2
22.3 EMV(a 1 ) = .4(55) + .1(43) + .3(29) + .2(15) = 38.0
EMV(a 2 ) = .4(26) + .1(38) + .3(43) + .2(51) = 37.3
843
22.4
22.5 a1 a2 a3
s1 0 15 21
s2 0 3 4
s3 20 5 0
22.6 EOL(a 1 ) = .2(0) + .6(0) + .2(20) = 4.0
EOL(a 2 ) = .2(15) + .6(3) + .2(5) = 5.8
EOL(a 3 ) = .2(21) + .6(4) + .2(0) = 6.6
The EOL decision is a 1 .
22.7a Produce
Demand a0 a1 a2 a3
s0 0 -3.00 -6.00 -9.00
s1 0 5.00 2.00 -1.00
s2 0 5.00 10.00 7.00
s3 0 5.00 10.00 15.00
844
b Produce
Demand a0 a1 a2 a3
s0 0 3.00 6.00 9.00
s1 5.00 0 3.00 6.00
s2 10.00 5.00 0 3.00
s3 15.00 10.00 5.00 0
22.8a EMV(a 0 ) = 0
EMV(a 1 ) = .25(-3.00) + .25(5.00) + .25(5.00) + .25(5.00) = 3.00
EMV(a 2 ) = . 25(-6.00) + .25(2.00) + .25(10.00) + .25(10.00) = 4.00
EMV(a 3 ) = . 25(-9.00) + .25(-1.00) + .25(7.00) + .25(15.00) = 3.00
845
EMV decision is a 2 (bake 2 cakes)
b EOL(a 0 ) = .25(0) + .25(5.00) + .25(10.00) + .25(15.00) = 7.50
EOL(a 1 ) = .25(3.00) + .25(0) + .25(5.00) + .25(10.00) = 4.50
EOL(a 2 ) = . 25(6.00) + .25(3.00) + .25(0) + .25(5.00) = 3.50
EOL(a 2 ) = . 25(9.00) + .25(6.00) + .25(3.00) + .25(0) = 4.50
EOL decision is a 2 (bake 2 cakes)
22.9 a 1 (flat fee) a 2 Pay per snowfall
s0 -40,000 0
s1 -40,000 -18,000
s2 -40,000 -36,000
s3 -40,000 -54,000
s4 -40,000 -72,000
22.10 EMV(a 1 ) = -40,000
EMV(a 2 ) = . 05(0) + .15(-18,000) + .30(-36,000) + .40(-54,000) +.10(-72,000) = -42,300
EMV decision is a 1
22.11a Payoff Table

a 100 a 200 a 300
s 100 12(100)-10(100) 12(100)-9(200)+6(100) 12(100)-8.50(300)+6(200)
= 200 =0 = -150
s 150 12(100)- 10(100) 12(150)-9(200)+6(50) 12(150)-8.50(300)+6(150)
= 200 = 300 = 150

s 200 12(100)-10(100) 12(200)-9(200) 12(200)-8.50(300)+6(100)
= 200 = 600 = 450

s 250 12(100)-10(100) 12(200)–9(200) 12(250)-8.50(300)+6(50)
= 200 = 600 = 750
846
b Opportunity Loss Table
a 100 a 200 a 300
s 100 0 200 350
s 150 100 0 150
s 200 400 0 150
s 250 550 150 0
22.12
EMV(a 100 ) = 200
EMV(a 200 ) = . 20(0) + .25(300) + .40(600) + .15(600) = 405
EMV(a 300 ) = . 20(-150) + .25(150) + .40(450) + .15(750) = 300
EMV decision is order 200 shirts.
847
22.13 P(s 0 ) = .607, P(s 1 ) = .303, P(s 2 ) = .076, P(s 3 ) = .012, P(s 4 ) = .002
Payoff Table
a0 a1 a2 a3
s0 0 -6,000 -12,000 -18,000
s1 0 7,000 1,000 -5,000
s2 0 7,000 14,000 8,000
s3 0 7,000 14,000 21,000
Opportunity Loss Table

a0 a1 a2 a3
s0 0 6,000 12,000 18,000
s1 7,000 0 6,000 12,000
s2 14,000 7,000 0 6,000
s3 21,000 14,000 7,000 0
22.14a EMV(Small) = .15(-220) + .55(-330) + .30(-440) = -346.5

EMV(Medium) = .15(-300) + .55(-320) + .30(-390) = -338.0
EMV(Large) = .15(-350) + .55(-350) + .30(-350) =-350.0
EMV decision: build a medium size plant; EMV*= -338.0
b Opportunity Loss Table

Small Medium Large
Low 0 80 130
Moderate 10 0 30
High 90 40 0
c EOL(Small) = .15(0) + .55(10) + .30(90) = 32.5

EOL(Medium) = .15(80) + .55(0) + .30(40) = 24.0
EOL(Large) = .15(130) + .55(30) + .30(0) = 36.0
EOL decision: build a medium size plant
848
22.15a P(s 10 ) = 9/90 = .10, P(s 11 ) = 18/90 = .20, P(s 12 ) = 36/90 = .40, P(s 13 ) = 27/90 = .30
Payoff Table
a 10 a 11 a 12 a 13
s 10 30 10(5)- 11(2)+2 10(5)-12(2)+3.50 10(5)-13(2)+4.50
= 30 = 29.50 = 28.50
s 11 30 11(5)-11(2) 11(5)-12(2)+2 11(5)-13(2)+3.50
= 33 = 33 = 32.50
s 12 30 11(5)-11(2) 12(5)-12(2) 12(5)- 13(2)+2
= 33 = 36 = 36
s 13 30 11(5)-11(2) 12(5)-12(2) 13(5)-13(2)
= 33 = 36 = 39
b EMV(a 10 ) = 30
EMV(a 11 ) = .10(30) + .20(33) + .40(33) + .30(33) = 32.70
EMV(a 12 ) = .10(29.50) + .20(33) + .40(36) + .30(36) = 34.75
EMV(a 13 ) = .10(28.50) + .20(32.50) + .40(36) + .30(39) = 35.45
EMV decision: buy 13 bushels
22.16 Payoff Table

Decision
Produce Don’t produce
Market share
5% -28 million 0
10% 2 million 0
15% 8 million 0
EMV(produce) = .15(-28 million) + .45(2 million) + .40 (8 million) = -.1 million

EMV (don’t produce) = 0
EMV decision: don’t produce
22.17 EPPI = .10(110) + .25(150) + .50(220) + .15(250) = 196

EMV(a 1 ) = .10(60) + .25(40) + .50(220) + .15(250) = 163.5
EMV(a 2 ) = .10(110) + .25(110) + .50(120) + .15(120) = 116.5
EMV(a 3 ) = .10(75) + .25(150) + .50(85) + .15(130) = 107
EVPI = EPPI – EMV* = 196 – 163.5 = 32.5
849
22.18 Opportunity Loss Table
a1 a2 a3
s1 50 0 35
s2 110 40 0
s3 0 100 135
s4 0 130 120
EOL(a 1 ) = .10(50) + .25(110) + .50(0) + .15(0) = 32.5
EOL(a 2 ) = .10(0) + .25(40) + .50(100) + .15(130) = 79.5
EOL(a 3 ) = .10(35) + .25(0) + .50(135) + .15(120) = 89
EOL* = 32.5
22.19 EPPI = .5(65) + .5(110) = 87.5
EMV(a 1 ) = .5(65) + .5(70) = 67.5
EMV(a 2 ) = .5(20) + .5(110) = 65.0
EMV(a 3 ) = .5(45) + .5(80) = 62.5
EMV(a 4 ) = .5(30) + .5(95) = 62.5
EVPI = EPPI – EMV* = 87.5 – 67.5 = 20
22.20 a EPPI = .75(65) + .25(110) = 76.25

EMV(a 1 ) = .75(65) + .25(70) = 66.25
EMV(a 2 ) = .75(20) + .25(110) = 42.5
EMV(a 3 ) = .75(45) + .25(80) = 53.75
EMV(a 4 ) = .75(30) + .25(95) = 46.25
EVPI = EPPI – EMV* = 76.25 – 66.25 = 10
b EPPI = .95(65) + .05(110) = 67.25

EMV(a 1 ) = .95(65) + .05(70) = 65.25
EMV(a 2 ) = .95(20) + .05(110) = 24.5
EMV(a 3 ) = .95(45) + .05(80) = 46.75
EMV(a 4 ) = .95(30) + .05(95) = 33.25
EVPI = EPPI – EMV* = 67.25 – 65.25 = 2
22.21 As the difference between the two prior probabilities increases EVPI decreases.
850
22.22 Posterior Probabilities for I 1
s j P(s j ) P(I 1 |s j ) P(s j and I 1 ) P(s j | I 1 )

__________________________________________________________________________
s1 .25 .40 (.25)(.40) = .10 .12/.20 = .500
s2 .40 .25 (.40)(.25) = .10 .10/.20 = .500
s3 .35 0 (.35)(0) = .0 0/.20 = 0
P(I 1 ) = .20
Posterior Probabilities for I 2

__________________________________________________________________________
s1 .25 .30 (.25)(.30) = .075 .075/.28 = .268
s2 .40 .25 (.40)(.25) = .10 .10/.28 = .357
s3 .35 30 (.35)(.30) = .105 .105/.28 = .375
P(I 2 ) = .28
__________________________________________________________________________
s1 .25 .20 (.25)(.20) = .05 .05/.29 = .172
s2 .40 .25 (.40)(.25) = .10 .10/.29 = .345
s3 .35 .40 (.35)(.40) = .14 .14/.29 = .483
P(I 3 ) = .29
__________________________________________________________________________
s1 .25 .10 (.25)(.10) = .025 .025/.23 = .109
s2 .40 .25 (.40)(.25) = .10 .10/.23 = .435
s3 .35 .30 (.35)(.30) = .105 .105/.23 = .456
P(I 4 ) = .23

__________________________________________________________________________
s1 .5 .98 (.5)(.98) = .49 .49/.515 = .951
s2 .5 .05 (.5)(.05) = .025 .025/.515 = .049
P(I 1 ) = .515
851
__________________________________________________________________________
s1 .5 .02 (.5)(.02) = .01 .01/.485 = .021
s2 .5 .95 (.5)(.95) = .475 .475/.485 = .979
P(I 2 ) = .485
22.24a Prior probabilities: EMV(a 1 ) = .5(10) + .5(22) = 16
EMV(a 2 ) = .5(18) + .5(19) = 18.5
EMV(a 3 ) = .5(23) + .5(15) = 19
EMV* = 19
I 1 : EMV(a 1 ) = .951(10) + .049(22) = 10.588
EMV(a 2 ) = .951(18) + .049(19) = 18.049
EMV(a 3 ) = .951(23) + .049(15) = 22.608
Optimal act: a 3
I 2 : EMV(a 1 ) = .021(10) + .979(22) = 21.748
EMV(a 2 ) = .021(18) + .979(19) = 18.979
EMV(a 3 ) = .021(23) + .979(15) = 15.168
Optimal act: a 1
b EMV` = .515(22.608) + .485(21.748) = 22.191

EVSI = EMV` - EMV* = 22.191 – 19 = 3.191
22.25 Prior probabilities: EMV(a 1 ) = .333(60) + .333(90) + .333(150) = 100
EMV(a 2 ) = 90
EMV* = 100

__________________________________________________________________________
s1 .333 .7 (.333)(.7) = .233 .233/.467 = .499
s2 .333 .5 (.333)(.5) = .167 .167/.467 = .358
s3 .333 .2 (.333)(.2) = .067 .067/.467 = .143
P(I 1 ) = .467
852
__________________________________________________________________________
s1 .333 .3 (.333)(.3) = .100 .100/.534 = .187
s2 .333 .5 (.333)(.5) = .167 .167/.534 = .313
s3 .333 .8 (.333)(.8) = .267 .267/.534 = .500
P(I 2 ) = .534
I 1 : EMV(a 1 ) = .499(60) + .358(90) + .143(150) = 83.61
EMV(a 2 ) = 90
I 2 : EMV(a 1 ) = .187(60) + .313(90) + .500(150) = 114.39
EMV(a 2 ) = 90
EMV` = .467(90) + .534(114.39) = 103.11

EVSI = EMV` - EMV* = 103.11 – 100 = 3.11
EMV(a 2 ) = 90
EMV* = 90

__________________________________________________________________________
s1 .5 .7 (.5)(.7) = .35 .35/.57 = .614
s2 .4 .5 (.4)(.5) = .20 .20/.57 = .351
s3 .1 .2 (.1)(.2) = .02 .02/.57 = .035
P(I 1 ) = .57
__________________________________________________________________________
s1 .5 .3 (.5)(.3) = .15 .15/.43 = .349
s2 .4 .5 (.4)(.5) = .20 .20/.43 = .465
s3 .1 .8 (.1)(.8) = .08 .08/.43 = .186
P(I 1 ) = .43
I 1 : EMV(a 1 ) = .614(60) + .351(90) + .035(150) = 73.68
EMV(a 2 ) = 90
I2: EMV(a 1 ) = .349(60) + .465(90) + .186(150) = 90.69
EMV(a 2 ) = 90
853
EMV` = .57(90) + .43(90.69) = 90.30
EVSI = EMV` - EMV* = 90.30 – 90 = .30
EMV(a 2 ) = 90
EMV* = 90

__________________________________________________________________________
s1 .90 .7 (.90)(.7) = .63 .63/.665 = .947
s2 .05 .5 (.05)(.5) = .025 .025/.665 = .038
s3 .05 .2 (.05)(.2) = .01 .01/.665 = .015
P(I 1 ) = .665
__________________________________________________________________________
s1 .90 .3 (.90)(.3) = .27 .27/.335 = .806
s2 .05 .5 (.05)(.5) = .025 .025/.335 = .075
s3 .05 .8 (.05)(.8) = .04 .04/.335 = .119
P(I 1 ) = .335
I 1 : EMV(a 1 ) = .947(60) + .038(90) + .015(150) = 62.49
EMV(a 2 ) = 90
I 2 : EMV(a 1 ) = .806(60) + .075(90) + .119(150) = 72.96
EMV(a 2 ) = 90
EMV` = .665(90) + .335(90) = 90

EVSI = EMV` - EMV* = 90 – 90 = 0
22.28 As the prior probabilities become more diverse EVSI decreases.
22.29
Payoff Table
Demand Purchase lot Don’t purchase lot
10,000 10,000(5)-125,000 = -75,000 0
30,000 30,000(5) – 125,000 = 25,000 0
50,000 50,000(5)-125,000 = 125,000 0
EMV(purchase) = .2(-75,000) + .5(25,000) + .3(125,000) = 35,000
854
EMV(don’t purchase) = 0
EPPI = .2(0) + .5(25,000) + .3(125,000) = 50,000
EVPI = EPPI – EMV* = 50,000 – 35,000 = 15,000
22.30 EMV* = 0
EPPI = .15(0) + .45(2 million) + .40(8 million) = 4.1 million
EVPI = EPPI – EMV* = 4.1 million – 0 = 4.1 million
22.31 Likelihood probabilities (binomial probabilities)

P(I | s 1 ) = P(x = 3, n= 25 | p = .05) = .0930
P(I | s 2 ) = P(x = 3, n= 25 | p = .10) = .2265
P(I | s 3 ) = P(x = 3, n= 25 | p = .15) = .2174
Posterior Probabilities
sj P(s j ) P(I | s j ) P(s j and I) P(s j | I)
__________________________________________________________________________
s1 .15 .0930 (.15)(.0930) = .0140 .0140/.2029 =
.0690
s2 .45 .2265 (.45)(.2265) = .1019 .1019/.2029 =
.5022
s3 .40 .2174 (.40)(.2174) = .0870 .0870/.2029 =
.4288
P(I) = .2029
EMV(produce) = .0690(-28 million) + .5022(2 million) + .4288(8 million) = 2.503 million

EMV (don’t produce) = 0
EMV decision: produce
22.32a Payoff Table

Market share Switch Don’t switch
5% 5(100,000) – 700,000 = -200,000 285,000
10% 10(100,000) – 700,000 = 300,000 285,000
20% 20(100,000)-700,000 = 1,300,000 285,000
b EMV(switch) = .4(-200,000) + .4(300,000) + .2(1,300,000) = 300,000
EMV(don’t switch) = 285,000
Optimal act: switch (EMV* = 300,000)
c EPPI = .4(285,000) + .4(300,000) + .2(1,300,000) = 494,000

EVPI = EPPI – EMV* = 494,000 – 300,000= 194,000
855
22.33 Payoff Table
Participating Households Proceed Don’t proceed
50,000 50(500) – 55,000 = -30,000 0
100,000 100(500) – 55,000 = -5,000 0
200,000 200(500) – 55,000 = 45,000 0
300,000 300(500) – 55,000 = 95,000 0
Likelihood probabilities (binomial probabilities)
P(I | s 1 ) = P(x = 3, n= 25 | p = .05) = .0930
P(I | s 2 ) = P(x = 3, n= 25 | p = .10) = .2265
P(I | s 3 ) = P(x = 3, n= 25 | p = .20) = .1358
P(I | s 4 ) = P(x = 3, n= 25 | p = .30) = .0243
__________________________________________________________________________
s1 .5 .0930 (.5)(.0930) = .0465 .0465/.1305 = .3563
s2 .3 .2265 (.3)(.2265) = .0680 .0680/.1305 = .5211
s3 .1 .1358 (.1)(.1358) = .0136 .0136/.1305 = .1042
s4 .1 .0243 (.1)(.0243) = .0024 .0024/.1305 = .0184
P(I) = .1305
EMV(proceed) = .3563(-30,000) + .5211(-5,000) + .1042(45,000) + .0184(95,000) = -6,858
EMV (don’t proceed = 0
EMV decision: don’t proceed
22.34 Likelihood probabilities (binomial probabilities)

P(I | s 1 ) = P(x = 12, n= 100 | p = .05) = .0028
P(I | s 2 ) = P(x = 12, n= 100 | p = .10) = .0988
P(I | s 3 ) = P(x = 12, n= 100 | p = .20) = .0128
P(I | s 4 ) = P(x = 12, n= 100 | p = .30) = .000013
__________________________________________________________________________
s1 .5 .0028 (.5)(.0028) = .0014 .0014/.0323 = .0433
s2 .3 .0988 (.3)(.0988) = .0296 .0296/.0323 = .9164
s3 .1 .0128 (.1)(.0128) = .0013 .0013/.0323 = .0402
s4 .1 .000013 (.1)(.000013) = .000001 .000001/.0323 = .000031
P(I) = .0323
856
EMV(proceed) = .0433(-30,000) + .9164(-5,000) + .0402(45,000) + .000031(95,000) = -4,069
EMV (don’t proceed = 0
EMV decision: don’t proceed

__________________________________________________________________________
s1 .15 .5 (.15)(.5) = .075 .075/.30 = .25
s2 .55 .3 (.55)(.3) = .165 .165/.30 = .55
s3 .30 .2 (.30)(.2) = .06 .06/.30 = .20
P(I 1 ) = .30

__________________________________________________________________________
s1 .15 .3 (.15)(.3) = .045 .045/.435 = .103
s2 .55 .6 (.55)(.6) = .33 .33/.435 = .759
s3 .30 .2 (.30)(.2) = .06 .06/.435 = .138
P(I 2 ) = .435

__________________________________________________________________________
s1 .15 .2 (.15)(.2) = .03 .03/.265 = .113
s2 .55 .1 (.55)(.1) = .055 .055/.265 = .208
s3 .30 .6 (.30)(.6) = .18 .18/.265 = .679
P(I 3 ) = .265
I 1 : EMV(a 1 ) = .25(-220) + .55(-330) + .20(-440) = -324.5
EMV(a 2 ) = .25(-300) + .55(-320) + .20(-390) = -329.0
EMV(a 3 ) = .251(-350) + .55(-350) + .20(-350) = -350
Optimal act: a 1
I 2 : EMV(a 1 ) = .103(-220) + .759(-330) + .138(-440) = -333.85
EMV(a 2 ) = .103(-300) + .759(-320) + .138(-390) = -327.59
EMV(a 3 ) = .103(-350) + .759(-350) + .138(-350) = -350
Optimal act: a 2
857
I 3 : EMV(a 1 ) = .113(-220) + .208(-330) + .679(-440) = -392.26
EMV(a 2 ) = .113(-300) + .208(-320) + .679(-390) = -365.28
EMV(a 3 ) = .113(-350) + .208(-350) + .679(-350) = -350
Optimal act: a 3
EMV` = .30(-324.5) + .435(-327.59) + .265(-350) = -332.60

EVSI = EMV` - EMV* = -332.60 – (-338) = 5.40
22.36
I 0 = neither person supports format change
I 1 = one person supports format change
I 2 = both people support format change
Likelihood probabilities P( I i | s j )
I0 I1 I2
5% .9025 .0950 .0025
10% .81 .18 .01
20% .64 .32 .04

__________________________________________________________________________
s1 .4 .9025 (.4)(.9025) = .361 .361/.813 = .444
s2 .4 .81 (.4)(.81) = .324 .324/.813 = .399
s3 .2 .64 (.2)(.64) = .128 .128/.813 = .157
P(I 0 ) = .813
__________________________________________________________________________
s1 .4 .0950 (.4)(.0950) = .038 .038/.174 = .218
s2 .4 .18 (.4)(.18) = .072 .072/.174 = .414
s3 .2 .32 (.2)(.32) = .064 .064/.174 = .368
P(I 1 ) = .174
__________________________________________________________________________
s1 .4 .0025 (.4)(.0025) = .001 .001/.013 = .077
s2 .4 .01 (.4)(.01) = .004 .004/.013 = .308
s3 .2 .04 (.2)(.04) = .008 .008/.013 = .615
P(I 2 ) = .013
858
I 1 : EMV(switch) = .444(-200,000) + .399(300,000) + .157(1,300,000) = 235,000

Optimal act: don’t switch
I 2 : EMV(switch) = .218(-200,000) + .414(300,000) + .368(1,300,000) = 559,000

Optimal act: switch
I 3 : EMV(switch) = .077(-200,000) + .308(300,000) + .615(1,300,000) = 876,500

Optimal act: switch
EMV` = .813(285,000) + .174(546,000) + .013(876,500) = 338,104
EVSI = EMV` - EMV* = 338,104 – 300,000 = 38,104
22.37
P(I | s 1 ) = P(x = 2, n= 25 | p = .05) = .2305
P(I | s 2 ) = P(x = 2, n= 25 | p = .10) = .2659
P(I | s 3 ) = P(x = 2, n= 25 | p = .20) = . 0708
Posterior Probabilities for I

sj P(s j ) P(I|s j ) P(s j and I) P(s j | I)
__________________________________________________________________________
s1 .4 .2305 (.4)(.2305) = .0922 .0922/.2127 = .4334
s2 .4 .2659 (.4)(.2659) = .1064 .1064/.2127 = .5000
s3 .2 .0708 (.2)(.0708) = .0142 .0142/.2127 = .0667
P(I) = .2127
EMV(switch) = .4334(-200,000) + .5000(300,000) + .0667(1,300,000) = 149,873

Optimal act: don’t switch
859
22.38a
Payoff Table
Demand Battery 1 Battery 2 Battery 3
50,000 20(50,000)-900,000 23(50,000)-1,150,000 25(50,000)-1,400,000
= 100,000 0 -150,000
100,000 20(100,000)-900,000 23(100,000)-1,150,000 25(100,000)-1,400,000
=1,100,000 1,150,000 1,100,000
150,000 20(150,000)-900,000 23(150,000)-1,150,000 25(150,000)-1,400,000
=2,100,000 2,300,000 2,350,000
b Opportunity Loss table

Demand Battery 1 Battery 2 Batter3
50,000 0 100,000 250,000
100,000 50,000 0 50,000
150,000 250,000 50,000 9
c EMV(Battery 1) = .3(100,000) + .3(1,100,000) + .4(2,100,000) = 1,200,000
EMV(Battery 2) = .3(0) + .3(1,150,000) + .4(2,300,000) = 1,265,000
EMV(Battery 3) = .3(-150,000) + .3(1,100,000) + .4(2,350,000) = 1,225,000
EMV decision: Battery 2
d EOL(Battery 2) = .3(100,000) + .3(0) + .4(50,000) = 50,000

EVPI = EOL* = 50,000
22.39 Payoff Table

Percentage change Change ad Don’t change
-2 -258,000 0
-1 -158,000 0
0 -58,000 0
1 42,000 0
2 142,000 0
EMV(Change ad) = -1(-258,000) + .1(-158,000) + .2(-58,000) + .3(42,000) + .3(142,000) = 2,000

EMV (don’t change) = 0.
Optimal decision: change ad
860
22.40
I 0 = person does not believe the ad
I 1 = person believes the ad
Likelihood probabilities P( I i | s j )
I0 I1
30% .70 .30
31% .69 .31
32% .68 .32
33% .67 .33
34% .66 .34

__________________________________________________________________________
s1 .1 .70 (.1)(.70) = .070 .070/.674 = .104
s2 .1 .69 (.1)(.69) = .069 .069/.674 = .102
s3 .2 .68 (.2)(.68) = .136 .136/.674 = .202
s4 .3 .67 (.3)(.67) = .201 .201/.674 = .298
s5 .3 .66 (.3)(.66) = .198 .198/.674 = .294
P(I 0 ) = .674
__________________________________________________________________________
s1 .1 .30 (.1)(.30) = .030 .030/.326 = .092
s2 .1 .31 (.1)(.31) = .031 .031/.326 = .095
s3 .2 .32 (.2)(.32) = .064 .064/.326 = .196
s4 .3 .33 (.3)(.33) = .099 .099/.326 = .304
s5 .3 .34 (.3)(.34) = .102 .102/.326 = .313
P(I 1 ) = .326
I 0 : EMV(Change ad) = .104(-258,000) + .102(-158,000) + .202(-58,000) + .298(42,000) +
.294(142,000)
= -400
Optimal decision: don’t change ad
I 1 : EMV(Change ad) = .092(-258,000) + .095(-158,000) + .196(-58,000) + .304(42,000) +
.313(142,000)
= 7,100
861
Optimal decision: change ad
EMV` = .674(0) + .326(7,100) = 2,315

EVSI = EMV` - EMV* = 2,315 – 2,000 = 315
22.41
P(I | s 1 ) = P(x = 1, n = 5 | p = .30) = .3602
P(I | s 2 ) = P(x = 1, n = 5 | p = .31) = .3513
P(I | s 3 ) = P(x = 1, n = 2 | p = .32) = . 3421
P(I | s 4 ) = P(x = 1, n = 2 | p = .33) = . 3325
P(I | s 5 ) = P(x = 1, n = 2 | p = .34) = . 3226
Posterior Probabilities for I

__________________________________________________________________________
s1 .1 .3602 (.1)(.3602) = .0360 .0360/.3361 = .1072
s2 .1 .3513 (.1)(.3513) = .0351 .0351/.3361 = .1045
s3 .2 .3421 (.2)(.3421) = .0684 .0684/.3361 = .2036
s4 .3 .3325 (.3)(.3325) = .0997 .0997/.3361 = .2968
s5 .3 .3226 (.3)(.3226) = .0968 .0968/.3361 = .2879
P(I) = .3361
EMV(Change ad) = .1072(-258,000) + .1045(-158,000) + .2036(-58,000) + .2968(42,000)

+ .2879(142,000) = -2,620
Optimal decision: don’t change ad
22.42
EMV(25 telephones) = 50,000
EMV(50 telephones) = .50(30,000) + .25(60,000) + .25(60,000) = 45,000
EMV(100 telephones) = .50(20,000) + .25(40,000) + .25(80,000) = 40,000
Optimal decision: 25 telephones (EMV* = 50,000)
I 1 = small number of calls
I 2 = medium number of calls
I 3 = large number of calls
862
Likelihood probabilities (Poisson distribution)
I1 I2 I3
µ=5 P(X < 8 | µ = 5) P(8 ≤ X < 17 | µ = 5) P(X ≥ 17 | µ = 5)
= .8667 = .1334 =0
µ = 10 P(X < 8 | µ = 10) P(8 ≤ X < 17 | µ = 10) P(X ≥ 17 | µ = 10)
= .2202 = .7527 = .0270

µ = 15 P(X < 8 | µ = 15) P(8 ≤ X < 17 | µ = 15) P(X ≥ 17 | µ = 15)
= .0180 = .6461 = .3359

__________________________________________________________________________
s1 .50 .8667 (.50)(.8667) = .4333 .4333/.4929 = .8792
s2 .25 .2202 (.25)(.2202) = .0551 .0551/.4929 = .1117
s3 .25 .0180 (.25)(.0180) = .0045 .0045/.4929 = .0091
P(I 1 ) = .4929

__________________________________________________________________________
s1 .50 .1334 (.50)(.1334) = .0667 .0667/.4164 = .1601
s2 .25 .7527 (.25)(.7527) = .1882 .1882/.4164 = .4519
s3 .25 .6461 (.25)(.6461) = .1615 .1615/.4164 = .3879
P(I 2 ) = .4164

__________________________________________________________________________
s1 .50 .0 (.50)(0) = 0 0/.0907 = 0
s2 .25 .0270 (.25)(.0270) = .0068 .0068/.0907 = .0745
s3 .25 .3359 (.25)(.3359) = .0840 .0840/.0907 = .9254
P(I 3 ) = .0907
I 1 : EMV(25 telephones) = 50,000
EMV(50 telephones) = .8792(30,000) + .1117(60,000) + .0091(60,000) = 33,624

EMV(100 telephones) = .8792(20,000) + .1117(40,000) + .0091(80,000) = 22,780
Optimal act: 25 telephones
863
EMV(50 telephones) = .1601(30,000) + .4519(60,000) + .3879(60,000) = 55,191

EMV(100 telephones) = .1601(20,000) + .4519(40,000) + .38791(80,000) = 52,310
EMV(50 telephones) = 0(30,000) + .0745(60,000) + .9254(60,000) = 60,000

EMV(100 telephones) = 0(20,000) + .0745(40,000) + .9254(80,000) = 77,012
EMV` = .4929(50,000) + .4164(55,191) + .0907(77,012) = 54,612
EVSI = EMV` - EMV* = 54,612 – 50,000 = 4,612
Because the value is greater than the cost ($4,000) Max should not sample. If he sees a small
number of calls install 25 telephones. If there is a medium number install 50 telephones. If there is
a large number of calls, install 100 telephones.
22.43a EMV(Model 101) = .2(20 million) + .4(100 million) + .4(210 million) = 128 million
EMV (Model 202) = .1(70 million) + .4(100 million) + .5(150 million) = 122 million
Optimal decision: Model 101
b Likelihood probabilities (binomial distribution) for Model 101

P(X =1, n = 10| p = .05) = .3151
P(X =1, n = 10| p = .10) = .3874
P(X =1, n = 10| p = .15) = .3474
Posterior Probabilities for Model 101

__________________________________________________________________________
s1 .2 .3151 (.2)(.3151) = .0630 .0630/.3570 = .1766
s2 .4 .3874 (.4)(.3874) = .1550 .1550/.3570 = .4341
s3 .4 .3474 (.4)(.3474) = .1390 .1390/.3570 = .3893
P(I) = .3570
EMV(Model 101) = .1766(20 million) + .4341(100 million) + .3893(210 million) = 128.7 million
Likelihood probabilities (binomial distribution) for Model 202

P(X =9, n = 20| p = .30) = .0654
P(X =9, n = 20| p = .40) = .1597
P(X =9, n = 20| p = .50) = .1602
864
Posterior Probabilities for Model 202
__________________________________________________________________________
s1 .1 .0654 (.1)(.0654) = .0065 .0065/.1505 = .0434
s2 .4 .1597 (.4)(.1597) = .0639 .0639/.1505 = .4245
s3 .5 .1602 (.5)(.1602) = .0801 .0801/.1505 = .5321
P(I) = .1505
EMV(Model 101) = .0434(70 million) + .4245(100 million) + .5321(150 million) = 125.3 million
Optimal decision: Model 101
22.44 EMV( Release in North America) = .5(33 million) + .3(12 million) + .2(-15 million) = 17.1
million
EMV(European distributor) = 12 million
Optimal decision: Release in North America
Posterior Probabilities for I 1 (Rave review)

__________________________________________________________________________
s1 .5 .8 (.5)(.8) = .40 .40/.63 = .635
s2 .3 .5 (.3)(.5) = .15 .15/.63 = .238
s3 .2 .4 (.2)(.4) = .08 .08/.63 = .127
P(I 1 ) = .63
EMV( Release in North America) = .635(33 million) + .238(12 million) + .127(-15 million) =21.9
million
Optimal decision: Release in North America
Posterior Probabilities for I 2 (lukewarm response)

__________________________________________________________________________
s1 .5 .1 (.5)(.1) = .05 .05/.20 = .25
s2 .3 .3 (.3)(.3) = .09 .09/.20 = .45
s3 .2 .3 (.2)(.3) = .06 .06/.20 = .30
P(I 2 ) = .20
865
EMV( Release in North America) = .25(33 million) + .45(12 million) + .30(-15 million) = 9.2
million
Optimal decision: Sell to European distributor
Posterior Probabilities for I 3 (poor response)

__________________________________________________________________________
s1 .5 .1 (.5)(.1) = .05 .05/.17 = .294
s2 .3 .2 (.3)(.2) = .06 .06/.17 = .353
s3 .2 .3 (.2)(.3) = .06 .06/.17 = .353
(I 3 ) = .17
EMV( Release in North America) = .294(33 million) + .353(12 million) + .353(-15 million) = 8.6
million
Optimal decision: Sell to European distributor.
EMV` = .63(21.9 million) + .20(12 million) + .17(12 million) = 18.2 million
EVSI = EMV` - EMV* = 18.2 million – 17.1 million = 1.1 million
Because EVSI is greater than the sampling cost (100,000) the studio executives should show the
movie to a random sample of North Americans. If the response is a rave review release the movie
in North America. If not sell it to Europe.
866

Statistics For Management and Economics, 9th Edition Keller ISM PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Statistics For Management and Economics, 9th Edition Keller ISM PDF

Hochgeladen von

Copyright:

Verfügbare Formate

How the Solutions Were Produced

acknowledge you with the discovery.

Chapters 10 through 19 and 21

all other tests.

Chapters 13, and Appendixes 13 to 17, and 19

column minus the variable in the second column.

Chapter 19 and Appendix 19

Excel produced all the solutions to these exercises.

All control charts were produced by Excel.

Solutions to these exercises were completed manually.

Statistics for Management and Economics

Wilfrid Laurier University

1.3a The political choices of the 25,000 registered voters

1.4a The complete production run

United Venezuela, 8% Brazil, 1%

Liquified Lubricants Asphalt and Other

Jet fuel Gasoline

Distillate fuel oil

Hydroelectric Biomass Other

Yard trimmings, Other, 3%

Textiles, 5% Glass, 5%

Recycled Plastic, Recycled Glass,

Discposed Recycled Other,

Disposed O ther, Recycled Glass ,

Pay medical Save Other

Pay down debt

Down payment Home

Career focus Location

2.29 a Newspaper Frequency Relative Frequency

New York Post

The Daily News and the Post dominate the market

2.30a Degree Frequency

2.32 a Software Frequency

Nine of out of 10 Americans classify themselves as middle or working class.

15,000 Some c ollege

Hydroelectric Biomass Other

New Zealand Energy Sources

The dominant source in Australia is coal. In New Zealand it is oil.

The two variables are related.

More than 40% rate the food as less than good.

There are considerable differences between the two countries.

More students disagree than agree.

c The histogram is unimodal and somewhat positively skewed.

c The number of pages is bimodal and slightly positively skewed.

d The ages are bimodal and positively skewed.

The histogram is bimodal.

c The histogram is positively skewed.

c The number of stores is bimodal and positively skewed.

The histogram is positively skewed.

The histogram is symmetric, unimodal, and bell shaped.

d The histogram is slightly positively skewed, unimodal, and not bell-shaped.

The histogram is unimodal and positively skewed.

3.16 a The histogram should contain 9 or 10 bins. We chose 10.

c The histogram is positively skewed.

The histogram is negatively skewed, bimodal, and not bell shaped.

c The scorecards appear to be relatively poor predictors.

c. and d. This scorecard is a much better predictor.

Many people work more than 40 hours per week.

After a rapid increase the numbers have leveled off.

Total health care expenditures are rising faster than inflation

The number of property crime decreased slowly over the 15 years.

GDP increased rapidly over the 29 year period.

The inflation-adjusted GDP grew at a moderate rate.

3-‐Yr share return