Page | 1

Chapter 1
Introduction to Statistics

LEARNING OBJECTIVES

The primary objective of chapter 1 is to introduce you to the world of

statistics, enabling you to:

1. Define statistics.

2. Be aware of a wide range of applications of statistics in business.

3. Differentiate between descriptive and inferential statistics.

4. Classify numbers by level of data and understand why doing so is

important.

CHAPTER TEACHING STRATEGY

In chapter 1 it is very important to motivate business students to study

statistics by presenting them with many applications of statistics in business.
The definition of statistics as a science dealing with the collection, analysis,
interpretation, and presentation of numerical data is a very good place to
start. Statistics is about dealing with data. Data are found in all areas of
business. This is a time to have the students brainstorm on the wide variety
of places in business where data are measured and gathered. It is important
to define statistics for students because they bring so many preconceptions
 Page | 2

of the meaning of the term. For this reason, several perceptions of the word
statistics are given in the chapter.

Chapter 1 sets up the paradigm of inferential statistics. The student

will understand that while there are many useful applications of descriptive
statistics in business, the strength of the application of statistics in the field of
business is through inferential statistics. From this notion, we will later
introduce probability, sampling, confidence intervals, and hypothesis testing.
The process involves taking a sample from the population, computing a
statistic on the sample data, and making an inference (decision or
conclusion) back to the population from which the sample has been drawn.

In chapter 1, levels of data measurement are emphasized. Too many

texts present data to the students with no comment or discussion of how the
data were gathered or the level of data measurement. In chapter 7, there is
a discussion of sampling techniques. However, in this chapter, four levels of
data are discussed. It is important for students to understand that the
statistician is often given data to analyze without input as to how it was
gathered or the type of measurement. It is incumbent upon statisticians and
researchers to ascertain the level of measurement that the data represent so
that appropriate techniques can be used in analysis. All techniques
presented in this text cannot be appropriately used to analyze all data.

CHAPTER OUTLINE

1.1 Statistics in Business

Marketing

Management

Finance

Economics

Management Information Systems

1.2 Basic Statistical Concepts

1.3 Data Measurement

Nominal Level
 Page | 3

Ordinal Level

Interval Level

Ratio Level

Comparison of the Four Levels of Data

Statistical Analysis Using the Computer: Excel and MINITAB

KEY TERMS

Census Ordinal Level Data

Descriptive Statistics Parameter

Inferential Statistics Parametric

Statistics

Interval Level Data Population

Metric Data Ratio Level Data

Nominal Level Data Sample

Non-metric Data Statistic

Nonparametric Statistics Statistics

SOLUTIONS TO PROBLEMS IN CHAPTER 1

1.1 Examples of data in functional areas:

 Page | 4

accounting - cost of goods, salary expense, depreciation, utility costs,

taxes, equipment inventory, etc.

finance - World bank bond rates, number of failed savings and loans,
measured risk of common stocks, stock dividends, foreign exchange
rate, liquidity rates for a single-family, etc.

human resources - salaries, size of engineering staff, years experience,

age of employees, years of education, etc.

marketing - number of units sold, dollar sales volume, forecast sales,

size of sales force, market share, measurement of consumer
motivation, measurement of consumer frustration, measurement of
brand preference, attitude measurement, measurement of consumer
risk, etc.

information systems - CPU time, size of memory, number of work

stations, storage capacity, percent of professionals who are connected
to a computer network, dollar assets of company computing, number
of “hits” on the Internet, time spent on the Internet per day,
percentage of people who use the Internet, retail dollars spent in e-
commerce, etc.

production - number of production runs per day, weight of a product;

assembly time, number of defects per run, temperature in the plant,
amount of inventory, turnaround time, etc.

management - measurement of union participation, measurement of

employer support, measurement of tendency to control, number of
subordinates reporting to a manager, measurement of leadership style,
etc.

1.2 Examples of data in business industries:

 Page | 5

manufacturing - size of punched hole, number of rejects, amount of

inventory, amount of production, number of production workers, etc.

insurance - number of claims per month, average amount of life

insurance per family head, life expectancy, cost of repairs for major
auto collision, average medical costs incurred for a single female over
45 years of age, etc.

travel - cost of airfare, number of miles traveled for ground transported

vacations, number of nights away from home, size of traveling party,
amount spent per day on besides lodging, etc.

retailing - inventory turnover ratio, sales volume, size of sales force,

number of competitors within 2 miles of retail outlet, area of store,
number of sales people, etc.

communications - cost per minute, number of phones per office, miles

of cable per customer headquarters, minutes per day of long distance
usage, number of operators, time between calls, etc.

computing - age of company hardware, cost of software, number of

CAD/CAM stations, age of computer operators, measure to evaluate
competing software packages, size of data base, etc.

agriculture - number of farms per county, farm income, number of

acres of corn per farm, wholesale price of a gallon of milk, number of
livestock, grain storage capacity, etc.

banking - size of deposit, number of failed banks, amount loaned to

foreign banks, number of tellers per drive-in facility, average amount
of withdrawal from automatic teller machine, federal reserve discount
rate, etc.
 Page | 6

healthcare - number of patients per physician per day, average cost of

hospital stay, average daily census of hospital, time spent waiting to
see a physician, patient satisfaction, number of blood tests done per
week.

1.3 Descriptive statistics in recorded music industry -

1) RCA total sales of compact discs this week, number of artists

under contract to a company at a given time.

2) Total dollars spent on advertising last month to promote an

album.

3) Number of units produced in a day.

4) Number of retail outlets selling the company's products.

Inferential statistics in recorded music industry -

1) Measure the amount spent per month on recorded music for a

few consumers then use that figure to infer the amount for the
population.

2) Determination of market share for rap music by randomly

selecting a sample of 500 purchasers of recorded music.

3) Determination of top ten single records by sampling the number

of requests at a few radio stations.
 Page | 7

4) Estimation of the average length of a single recording by taking

a sample of records and measuring them.

The difference between descriptive and inferential statistics lies mainly

in the usage of the data. These descriptive examples all gather data
from every item in the population about which the description is being
made. For example, RCA measures the sales on all its compact discs
for a week and reports the total.

In each of the inferential statistics examples, a sample of the

population is taken and the population value is estimated or inferred
from the sample. For example, it may be practically impossible to
determine the proportion of buyers who prefer rap music. However, a
random sample of buyers can be contacted and interviewed for music
preference. The results can be inferred to population market share.

1.4 Descriptive statistics in manufacturing batteries to make better

decisions -

1) Total number of worker hours per plant per week - help

management understand labor costs, work allocation,
productivity, etc.

2) Company sales volume of batteries in a year - help management

decide if the product is profitable, how much to advertise in the
coming year, compare to costs to determine profitability.

3) Total amount of sulfuric acid purchased per month for use in

battery production. - can be used by management to study
wasted inventory, scrap, etc.

Inferential Statistics in manufacturing batteries to make decisions -

 Page | 8

1) Take a sample of batteries and test them to determine the

average shelf life - use the sample average to reach conclusions
about all batteries of this type. Management can then make
labeling and advertising claims. They can compare these figures
to the shelf-life of competing batteries.

2) Take a sample of battery consumers and determine how many

batteries they purchase per year. Infer to the entire population -
management can use this information to estimate market
potential and penetration.

3) Interview a random sample of production workers to determine

attitude towards company management - management can use
this survey result to ascertain employee morale and to direct
efforts towards creating a more positive working environment
which, hopefully, results in greater productivity.

1.5 a) ratio

b) ratio

c) ordinal

d) nominal

e) ratio
f) ratio

g) nominal

h) ratio

1.6 a) ordinal

b) ratio
c) nominal

d) ratio
e) interval

f) interval
 Page | 9

g) nominal
h) ordinal

1.7 a) The population for this study is the 900 electric contractors who
purchased Rathburn wire.

b) The sample is the randomly chosen group of thirty-five

contractors.

c) The statistic is the average satisfaction score for the sample of

thirty-five contractors.

d) The parameter is the average satisfaction score for all 900

electric contractors in the population.
 P a g e | 10

Chapter 2
Charts and Graphs

LEARNING OBJECTIVES

The overall objective of chapter 2 is for you to master several techniques for
summarizing and depicting data, thereby enabling you to:

1. Recognize the difference between grouped and ungrouped data.

2. Construct a frequency distribution.

3. Construct a histogram, a frequency polygon, an ogive, a pie chart, a

stem and leaf plot, a Pareto chart, and a scatter plot.

CHAPTER TEACHING STRATEGY

Chapter 1 brought to the attention of students the wide variety and

amount of data available in the world of business. In chapter 2, we confront
the problem of trying to begin to summarize and present the data in a
 P a g e | 11

meaningful manner. One mechanism for data summarization is the

frequency distribution, which is essentially a way of organizing ungrouped or
raw data into grouped data. It is important to realize that there is
considerable art involved in constructing a frequency distribution. There are
nearly as many possible frequency distributions for a problem as there are
students in a class. Students should begin to think about the receiver or user
of their statistical product. For example, what class widths and class
endpoints would be most familiar and meaningful to the end user of the
distribution? How can the data best be communicated and summarized using
the frequency distribution?

The second part of chapter 2 presents various ways to depict data

using graphs. The student should view these graphical techniques as tools
for use in communicating characteristics of the data in an effective manner.
Most business students will have some type of management opportunity in
their field before their career ends. The ability to make effective
presentations and communicate their ideas in succinct, clear ways is an
asset. Through the use of graphics packages and such techniques as
frequency polygons, ogives, histograms, and pie charts, the manager can
enhance his/her personal image as a communicator and decision-maker. In
addition, the manager can emphasize that the final product (the frequency
polygon, etc.) is just the beginning. Students should be encouraged to study
the graphical output to recognize business trends, highs, lows, etc. and
realize that the ultimate goal for these tools is their usage in decision making.

CHAPTER OUTLINE

2.1 Frequency Distributions

Class Midpoint

Relative Frequency

Cumulative Frequency

2.2 Graphic Depiction of Data

 P a g e | 12

Histograms

Frequency Polygons

Ogives

Pie Charts

Stem and Leaf Plots

Pareto Charts

2.3 Graphical Depiction of Two-Variable Numerical Data: Scatter Plots

KEY TERMS

Class Mark Pareto Chart

Class Midpoint Pie Chart

Cumulative Frequency Range

Frequency Distribution Relative Frequency

Frequency Polygon Scatter Plot

Grouped Data Stem and Leaf Plot

Histogram Ungrouped Data

Ogive
 P a g e | 13

SOLUTIONS TO PROBLEMS IN CHAPTER 2

2.1

a) One possible 5 class frequency distribution:

Class Interval Frequency

10 - under 25 9

25 - under 40 13

40 - under 55 11

55 - under 70 9

70 - under 85 8

50

b) One possible 10 class frequency distribution:

Class Interval Frequency

10 - under 18 7

18 - under 26 3

26 - under 34 5

34 - under 42 9

42 - under 50 7

50 - under 58 3

58 - under 66 6

66 - under 74 4

74 - under 82 4
 P a g e | 14

82 - under 90 2

c) The ten class frequency distribution gives a more detailed

breakdown of temperatures, pointing out the smaller frequencies for
the higher temperature intervals. The five class distribution
collapses the intervals into broader classes making it appear that
there are nearly equal frequencies in each class.
 P a g e | 15

2.2 One possible frequency distribution is the one below with 12 classes
and class intervals of 2.

Class Interval Frequency

39 - under 41 2

41 - under 43 1

43 - under 45 5

45 - under 47 10

47 - under 49 18

49 - under 51 13

51 - under 53 15

53 - under 55 15

55 - under 57 7

57 - under 59 9

59 - under 61 4

61 – under 63 1

The distribution reveals that only 13 of the 100 boxes of raisins contain
50 ± 1 raisin (49 -under 51). However, 71 of the 100 boxes of raisins
contain between 45 and 55 raisins. It shows that there are five boxes
that have 9 or more extra raisins (59-61 and 61-63) and two boxes that
have 9-11 less raisins (39-41) than the boxes are supposed to contain.

2.3

Class Class Relative Cumulative

Interval Frequency Midpoint Frequency Frequency

 P a g e | 16

0-5 6 2.5 6/86 = .0698 6

5 - 10 8 7.5 .0930 14

10 - 15 17 12.5 .1977 31

15 - 20 23 17.5 .2674 54

20 - 25 18 22.5 .2093 72

25 - 30 10 27.5 .1163 82

30 - 35 4 32.5 .0465 86

TOTAL 86 1.0000

The relative frequency tells us that it is most probable that a customer

is in the

15 - 20 category (.2674). Over two thirds (.6744) of the customers are

between 10

and 25 years of age.

 P a g e | 17

2.4

Class Class Relative Cumulative

Interval Frequency Midpoint Frequency

Frequency

0-2 218 1 .436 218

2-4 207 3 .414 425

4-6 56 5 .112 481

6-8 11 7 .022 492

8-10 8 9 .016 500

TOTAL 500 1.000

2.5 Some examples of cumulative frequencies in business:

sales for the fiscal year,

costs for the fiscal year,

spending for the fiscal year,

inventory build-up,

accumulation of workers during a hiring buildup,

production output over a time period.

2.6 Histogram:
P a g e | 18
 P a g e | 19

Frequency Polygon:

2.7 Histogram:
P a g e | 20
 P a g e | 21

Frequency Polygon:

2.8 Ogive:
P a g e | 22
 P a g e | 23

2.9 STEM LEAF

21 2, 8, 8, 9

22 0, 1, 2, 4, 6, 6, 7, 9, 9

23 0, 0, 4, 5, 8, 8, 9, 9, 9, 9

24 0, 0, 3, 6, 9, 9, 9

25 0, 3, 4, 5, 5, 7, 7, 8, 9

26 0, 1, 1, 2, 3, 3, 5, 6

27 0, 1, 3

2.10 Firm Proportion Degrees

Caterpillar .372 134

Deere .246 89
Illinois Too Works .144 52

Eaton .121 44

American Standard .117 42

TOTAL 1.000 361

Pie Chart:
 P a g e | 24

Annual Sales

American Standard; 12%

Eaton; 12% Caterpillar; 37%

Illinois Tool Works; 14%

Deere; 25%
 P a g e | 25

2. 11 Company Proportion Degrees

Southwest .202 73

Delta .198 71

American .181 65

United .140 50

Northwest .108 39

US Airways .094 34

Continental .078 28

TOTAL 1.001 360

Pie Chart:

Em planem ents

Continental; 8%
Southw est; 20%
US Airw ays; 9%

Northw est; 11%

Delta; 20%

United; 14%

American; 18%
 P a g e | 26

2.12 Brand Proportion Degrees

Pfizer .289 104

Johnson & Johnson .259 93

Merck .125 45

Bristol-Myers Squibb .120 43

Abbott Laboratories .112 40

Wyeth .095 34

TOTAL 1.000 359

Pie Chart:

Pharm ace utical Sales

Wyeth; 9%

Abbott Lab.; 11% Pfizer; 29%

Bristol-Myers Squibb; 12%

Merck; 13%
Johnson & Johnson; 26%
P a g e | 27
 P a g e | 28

2.13 STEM LEAF

1 3, 6, 7, 7, 7, 9, 9, 9

2 0, 3, 3, 5, 7, 8, 9, 9

3 2, 3, 4, 5, 7, 8, 8

4 1, 4, 5, 6, 6, 7, 7, 8, 8, 9

5 0, 1, 2, 2, 7, 8, 9

6 0, 1, 4, 5, 6, 7, 9

7 0, 7

8 0

The stem and leaf plot shows that the number of passengers per flight
were

relatively evenly distributed between the high teens through the

sixties. Rarely

was there a flight with at least 70 passengers. The category of 40's

contained the

most flights (10).

2.14 Complaint Number % of Total

Busy Signal 420 56.45

Too long a Wait 184 24.73

Could not get through 85 11.42

Got Disconnected 37 4.97

Transferred to the Wrong Person 10 1.34

 P a g e | 29

Poor Connection 8 1.08

Total 744 99.99

Customer Complaints
800
700
100
600 80

Percent
500
Count

400
60
300 40
200
100
20
0 0
C1

Count 420 184 85 37 10 8

Percent 56.5 24.7 11.4 5.0 1.3 1.1
Cum % 56.5 81.2 92.6 97.6 98.9 100.0
 P a g e | 30

2.15

3500

3000

2500

2000

Industrial Products
1500

1000

500

0
3000 3500 4000 4500 5000 5500 6000 6500

Hum an Food
 P a g e | 31

2.16

180

160

140

120

100

Sales
80

60

40

20

0
1 2 3 4 5 6 7 8 9 10 11

Adve rtising

2.17 Class Interval Frequencies

16 - under 23 6

23 - under 30 9

30 - under 37 4

37 - under 44 4

44 - under 51 4

51 - under 58 3

TOTAL 30
 P a g e | 32

2.18

Class Interval Frequency Midpoint Rel. Freq. Cum. Freq.

20 - under 25 17 22.5 .207 17

25 - under 30 20 27.5 .244 37

30 - under 35 16 32.5 .195 53

35 - under 40 15 37.5 .183 68

40 - under 45 8 42.5 .098 76

45 - under 50 6 47.5 .073 82

 P a g e | 33

2.19 Class Interval Frequencies

50 - under 60 13

60 - under 70 27

70 - under 80 43

80 - under 90 31

90 - under 100 9

TOTAL 123

Histogram:

Frequency Polygon:
 P a g e | 34

Ogive:

2.20 Label Value Proportion Degrees

A 55 .180 65
B 121 .397 143
 P a g e | 35

C 83 .272 98

D 46 .151 54

TOTAL 305 1.000 360

Pie Chart:

D; 15% A; 18%

C; 27%

B; 40%

2.21 STEM LEAF

28 4, 6, 9

29 0, 4, 8

30 1, 6, 8, 9

31 1, 2, 4, 6, 7, 7

32 4, 4, 6

33 5
 P a g e | 36

2.22 Problem Frequency Percent of Total

1 673 26.96

2 29 1.16

3 108 4.33

4 379 15.18

5 73 2.92

6 564 22.60

7 12 0.48

8 402 16.11

9 54 2.16

10 202 8.09

2496

Pareto Chart:
 P a g e | 37

Problems

2500 100

2000 80

1500 60

Percent
Count

1000 40

500 20

0 0
C5 1 6 8 4 10 3 5 9 2 Other
Count 673 564 402 379 202 108 73 54 29 12
Percent 27.0 22.6 16.1 15.2 8.1 4.3 2.9 2.2 1.2 0.5
Cum % 27.0 49.6 65.7 80.8 88.9 93.3 96.2 98.4 99.5 100.0

2.23
 P a g e | 38

16

14

12

10

8
Y

0
4 6 8 10 12 14 16 18

2.24 Whitcomb Company

Class Interval Frequency

32 - under 37 1

37 - under 42 4

42 - under 47 12

47 - under 52 11

52 - under 57 14

57 - under 62 5

62 - under 67 2
 P a g e | 39

67 - under 72 1

TOTAL 50
 P a g e | 40

2.25

Class Class Relative Cumulative

Interval Frequency Midpoint Frequency Frequency

20 – 25 8 22.5 8/53 = .1509 8

25 – 30 6 27.5 .1132 14

30 – 35 5 32.5 .0943 19

35 – 40 12 37.5 .2264 31

40 – 45 15 42.5 .2830 46

45 – 50 7 47.5 .1321 53

TOTAL 53 .9999

2.26 Frequency Distribution:

Class Interval Frequency

10 - under 20 2

20 - under 30 3

30 - under 40 9

40 - under 50 7

50 - under 60 12

60 - under 70 9

70 - under 80 6

80 - under 90 2

50
 P a g e | 41

Histogram:
 P a g e | 42

Frequency Polygon:

The normal distribution appears to peak near the center and diminish
towards the

end intervals.

2.27

a. Histogram and a Frequency Polygon for Problem 2.25

Class Cumulative

Interval Frequency Frequency

 P a g e | 43

20 – 25 8 8

25 – 30 6 14

30 – 35 5 19

35 – 40 12 31

40 – 45 15 46

45 – 50 7 53

TOTAL 53
 P a g e | 44

Histogram:

Frequency Polygon:
P a g e | 45
 P a g e | 46

b. Ogive:

2.28 Cumulative
 P a g e | 47

Asking Price Frequency Frequency

$ 80,000 - under $ 100,000 21 21

$ 100,000 - under $ 120,000 27 48

$ 120,000 - under $ 140,000 18 66

$ 140,000 - under $ 160,000 11 77

$ 160,000 - under $ 180,000 6 83

$ 180,000 - under $ 200,000 3 86

86
 P a g e | 48

Histogram:

30

25

20

15
Frequency

10

0
90000 110000 130000 150000 170000 190000

Class Midpoints

Frequency Polygon:
 P a g e | 49

30

25

20

15
Frequency

10

0
90000 110000 130000 150000 170000 190000

Class Midpoints

Ogive:
 P a g e | 50

100

90

80

70

60

50
Frequencies

40

30

20

10

0
80000 100000 120000 140000 160000 180000 200000

Class Endpoints

2.29 Amount Spent Cumulative

on Prenatal Care Frequency Frequency

$ 0 - under $100 3 3

$100 - under $200 6 9

$200 - under $300 12 21

$300 - under $400 19 40

$400 - under $500 11 51

$500 - under $600 6 57

57
P a g e | 51
 P a g e | 52

Histogram:

Frequency Polygon:
P a g e | 53
 P a g e | 54

Ogive:

2.30 Cumulative

Price Frequency Frequency

$1.75 - under $1.90 9 9

$1.90 - under $2.05 14 23

$2.05 - under $2.20 17 40

 P a g e | 55

$2.20 - under $2.35 16 56

$2.35 - under $2.50 18 74

$2.50 - under $2.65 8 82

$2.65 - under $2.80 5 87

87
 P a g e | 56

Histogram:

Frequency Polygon:
 P a g e | 57

Ogive:
 P a g e | 58

2.31 Genre Albums Sold Proportion Degrees

R&B 146.4 .29 104

Alternative 102.6 .21 76

Rap 73.7 .15 54

Country 64.5 .13 47

Soundtrack 56.4 .11 40

Metal 26.6 .05 18

Classical 14.8 .03 11

Latin 14.5 .03 11

TOTAL 1.00 361

Pie Chart:
 P a g e | 59

Classical; 3% Latin; 3%

Metal ; 5%
R&B; 29%
Soundtrack; 11%

Country; 13%

Alternative; 21%
Rap; 15%
 P a g e | 60

2.32

700

600

500

400

M anufactured Goods
300

200

100

0
0 5 10 15 20 25 30 35

Agricultural Products
 P a g e | 61

2.33

Industry Total Release Proportion Degrees

Chemicals 737,100,000 .366 132

Primary metals 566,400,000 .281 103

Paper 229,900,000 .114 41

Plastics & Rubber 109,700,000 .054 19

Transportation

Equipment 102,500,000 .051 18

Food 89,300,000 .044 16

Fabricated Metals 85,900,000 .043 15

Petroleum 63,300,000 .031 11

Electrical

Equipment 29,100,000 .014 5

TOTAL 0.998 360

Pie Chart:
 P a g e | 62

Petro.; 3% Elec.; 1%
Fab. Metals; 4%
Food; 4%

Trans. Equip.; 5%
Chem.; 37%
Plas. & Rubber; 5%

Paper; 11%

Prim. Metals; 28%

 P a g e | 63

2.34

Causes of Poor Quality Bottles

500 100

400 80

Percent
300 60
Count

200 40

100 20

0 0
C1 Fault in plastic Thickness Broken handle Labeling Discoloration
Count 221 117 86 44 32
Percent 44.2 23.4 17.2 8.8 6.4
Cum % 44.2 67.6 84.8 93.6 100.0

2.35 STEM LEAF

42 12, 16, 24, 32, 99, 99

43 04, 28, 39, 46, 61, 88

44 20, 40, 59
 P a g e | 64

45 12

46 53, 54

47 30, 34, 58

48 22, 34, 66, 78

49 63

50 48, 49, 90

51 66

52 21, 54, 57, 63, 91

53 38, 66, 66

54 31, 78

55 56

56 69

57 37, 50

58 31, 32, 58, 73

59 19, 23

2.36 STEM LEAF

22 00, 68

23 01, 37, 44, 75

24 05, 37, 48, 60, 68

25 24, 55

26 02, 56, 70, 77

27 42, 60, 64

28 14, 30

29 22, 61, 75, 76, 90, 96

 P a g e | 65

30 02, 10

2.37 The distribution of household income is bell-shaped with an average of

about

$ 90,000 and a range of from $ 30,000 to $ 140,000.

2.38 Family practice is most prevalent with about 20% with pediatrics next at
slightly

less. A virtual tie exists between ob/gyn, general surgery,

anesthesiology, and

psychiatry at about 14% each.

2.39 The fewest number of audits is 12 and the most is 42. More companies
(8)

performed 27 audits than any other number. Thirty-five companies

performed

between 12 and 19 audits. Only 7 companies performed 40 or more

audits.

2.40 There were relatively constant sales from January through August ($4 to
6 million).
 P a g e | 66

Each month from September through December sales increased with

December

having the sharpest increase ($15 million in sales in December).

Chapter 3
Descriptive Statistics

LEARNING OBJECTIVES

The focus of Chapter 3 is on the use of statistical techniques to describe data,

thereby enabling you to:

1. Distinguish between measures of central tendency, measures of

variability, and measures of shape.

2. Understand conceptually the meanings of mean, median, mode,

quartile, percentile, and range.

3. Compute mean, median, mode, percentile, quartile, range, variance,

standard deviation, and mean absolute deviation on ungrouped data.

4. Differentiate between sample and population variance and standard

deviation.

5. Understand the meaning of standard deviation as it is applied using

the empirical rule and Chebyshev’s theorem.

6. Compute the mean, median, standard deviation, and variance on

grouped data.

7. Understand box and whisker plots, skewness, and kurtosis.

8. Compute a coefficient of correlation and interpret it.

 P a g e | 67

CHAPTER TEACHING STRATEGY

In chapter 2, the students learned how to summarize data by

constructing frequency distributions (grouping data) and by using graphical
depictions. Much of the time, statisticians need to describe data by using
single numerical measures. Chapter 3 presents a cadre of statistical
measures for describing numerically sets of data.

It can be emphasized in this chapter that there are at least two major
dimensions along which data can be described. One is the measure of
central tendency with which statisticians attempt to describe the more
central portions of the data. Included here are the mean, median, mode,
percentiles, and quartiles. It is important to establish that the median is a
useful device for reporting some business data, such as income and housing
costs, because it tends to ignore the extremes. On the other hand, the mean
utilizes every number of a data set in its computation. This makes the mean
an attractive tool in statistical analysis.

A second major group of descriptive statistical techniques are the

measures of variability. Students can understand that a measure of central
tendency is often not enough to fully describe data, often giving information
only about the center of the distribution or key milestones of the distribution.
A measure of variability helps the researcher get a handle on the spread of
the data. An attempt is made in this text to communicate to the student that
through the use of the empirical rule and/or Chebyshev’s Theorem, students
can better understand the meaning of a standard deviation. The empirical
rule will be referred to quite often throughout the course; and therefore, it is
important to emphasize it as a rule of thumb. For example, in discussing
control charts in chapter 18, the upper and lower control limits are
established by using the range of + 3 standard deviations of the statistic as
limits within which 99.7% of the data values should fall if a process is in
control.

In this section of chapter 3, z scores are presented mainly to bridge the

gap between the discussion of means and standard deviations in chapter 3
and the normal curve of chapter 6. One application of the standard deviation
in business is the use of it as a measure of risk in the financial world. For
 P a g e | 68

example, in tracking the price of a stock over a period of time, a financial

analyst might determine that the larger the standard deviation, the greater
the risk (because of “swings” in the price). However, because the size of a
standard deviation is a function of the mean and a coefficient of variation
conveys the size of a standard deviation relative to its mean, other financial
researchers prefer the coefficient of variation as a measure of the risk. That
is, it can be argued that a coefficient of variation takes into account the size
of the mean (in the case of a stock, the investment) in determining the
amount of risk as measured by a standard deviation.

It should be emphasized that the calculation of measures of central

tendency and variability for grouped data is different than for ungrouped or
raw data. While the principles are the same for the two types of data,
implementation of the formulas is different. Computations of statistics from
grouped data are based on class midpoints rather than raw values; and for
this reason, students should be cautioned that group statistics are often just
approximations.

Measures of shape are useful in helping the researcher describe a

distribution of data. The Pearsonian coefficient of skewness is a handy tool for
ascertaining the degree of skewness in the distribution. Box and Whisker
plots can be used to determine the presence of skewness in a distribution and
to locate outliers. The coefficient of correlation is introduced here instead of
chapter 14 (regression chapter) so that the student can begin to think about
two-variable relationships and analyses and view a correlation coefficient as a
descriptive statistic. In addition, when the student studies simple regression
in chapter 14, there will be a foundation upon which to build. All in all,
chapter 3 is quite important because it presents some of the building blocks
for many of the later chapters.
 P a g e | 69

CHAPTER OUTLINE

3.1 Measures of Central Tendency: Ungrouped Data

Mode
Median

Mean

Percentiles

Quartiles

3.2 Measures of Variability - Ungrouped Data

Range

Interquartile Range

Mean Absolute Deviation, Variance, and Standard Deviation

Mean Absolute Deviation

Variance

Standard Deviation

Meaning of Standard Deviation

Empirical Rule

Chebyshev’s Theorem

Population Versus Sample Variance and Standard

Deviation

Computational Formulas for Variance and Standard

Deviation

Z Scores

Coefficient of Variation
 P a g e | 70

3.3 Measures of Central Tendency and Variability - Grouped Data

Measures of Central Tendency

Mean

Mode

Measures of Variability

3.4 Measures of Shape

Skewness

Skewness and the Relationship of the Mean, Median, and

Mode

Coefficient of Skewness

Kurtosis

Box and Whisker Plot

3.5 Measures of Association

Correlation

3.6 Descriptive Statistics on the Computer

KEY TERMS

Arithmetic Mean Measures of Shape

Bimodal Measures of Variability

 P a g e | 71

Box and Whisker Plot Median

Chebyshev’s Theorem Mesokurtic

Coefficient of Correlation (r) Mode

Coefficient of Skewness Multimodal

Coefficient of Variation (CV) Percentiles

Correlation Platykurtic

Deviation from the Mean Quartiles

Empirical Rule Range

Interquartile Range Skewness

Kurtosis Standard Deviation

Leptokurtic Sum of Squares of x

Mean Absolute Deviation (MAD) Variance

Measures of Central Tendency z Score

SOLUTIONS TO PROBLEMS IN CHAPTER 3

3.1 Mode

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

The mode = 4
 P a g e | 72

4 is the most frequently occurring value

3.2 Median for values in 3.1

Arrange in ascending order:

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

There are 15 terms.

Since there are an odd number of terms, the median is the middle
number.

The median = 4

Using the formula, the median is located

n 1 151
2 2
th
at the term = = 8th term

The 8th term =

4

3.3 Median

Arrange terms in ascending order:

073, 167, 199, 213, 243, 345, 444, 524, 609, 682

There are 10 terms.

Since there are an even number of terms, the median is the average of
the

middle two terms:

(2 4 3 3 4 5) 5 8 8

2 2
Median = = 294
 P a g e | 73

n 1
2
th
Using the formula, the median is located at the term

10  1 11

2 2
n = 10 therefore = 5.5th term.

The median is located halfway between the 5 th and 6th terms.

5th term = 243 6th term = 345

Halfway between 243 and 345 is the median = 294

3.4 Mean

17.3

44.5 µ = x/N = (333.6)/8 = 41.7

31.6

40.0

x
52.8 = x/n = (333.6)/8 = 41.7

38.8

30.1

78.5 (It is not stated in the problem whether the

x = 333.6 data represent as population or a sample).

3.5 Mean

-2
 P a g e | 74

5 µ = x/N = -12/12 = -1

-3

x
-6 = x/n = -12/12 = -1

-7

-4

-5

-8 (It is not stated in the problem whether the

x = -12 data represent a population or a sample).

3.6 Rearranging the data into ascending order:

11, 13, 16, 17, 18, 19, 20, 25, 27, 28, 29, 30, 32, 33, 34

35
i (15)  5.25
100

P35 is located at the 5 + 1 = 6th term, P35 = 19

55
i (15)  8.25
100

P55 is located at the 8 + 1 = 9th term, P55 = 27

 P a g e | 75

25
i (15)  3.75
100
Q1 = P25 but

Q1 = P25 is located at the 3 + 1 = 4th term, Q1 = 17

th
 15  1 
   8th term
 2 
Q2 = Median but: The median is located at the

Q2 = 25

75
i (15)  11.25
100
Q3 = P75 but

Q3 = P75 is located at the 11 + 1 = 12th term, Q3 = 30

 P a g e | 76

3.7 Rearranging the data in ascending order:

80, 94, 97, 105, 107, 112, 116, 116, 118, 119, 120, 127,

128, 138, 138, 139, 142, 143, 144, 145, 150, 162, 171, 172

n = 24

20
i ( 24)  4.8
100
For P20:

Thus, P20 is located at the 4 + 1 = 5th term and P20 = 107

47
i (24)  11.28
100
For P47:

Thus, P47 is located at the 11 + 1 = 12th term and P47 = 127

83
i (24)  19.92
100
For P83:

Thus, P83 is located at the 19 + 1 = 20th term and P83 = 145

Q1 = P25
 P a g e | 77

25
i (24)  6
100
For P25:

Thus, Q1 is located at the 6.5th term and Q1 = (112 + 116)/ 2 = 114

Q2 = Median

th
 24  1 
   12.5th term
 2 
The median is located at the:

Thus, Q2 = (127 + 128)/ 2 = 127.5

Q3 = P75

75
i (24)  18
100
For P75:

Thus, Q3 is located at the 18.5th term and Q3 = (143 + 144)/ 2 =

143.5

 x  18,245  1216.33
N 15
3.8 Mean = The mean is 1216.33.

th
 15  1 
 
 2 
The median is located at the = 8th term

Median = 1,233
 P a g e | 78

Q2 = Median = 1,233

63
i (15)  9.45
100
For P63,

P63 is located at the 9 + 1 = 10th term, P63 = 1,277

29
i (15)  4.35
100
For P29,

P29 is located at the 4 + 1 = 5th term, P29 = 1,119

th
 12  1 
   6.5 th
 2 
3.9 The median is located at the position

The median = (3.41 + 4.63)/2 = 4.02

75
i (12)  9
100
For Q3 = P75:

P75 is located halfway between the 9th and 10th terms.

 P a g e | 79

Q3 = P75 Q3 = (5.70 + 7.88)/ 2 = 6.79

20
i (12)  2.4
100
For P20:

P20 is located at the 3rd term P20 = 2.12

60
i (12)  7.2
100
For P60:

P60 is located at the 8th term P60 = 5.10

80
i (12)  9.6
100
For P80:

P80 is located at the 10th term P80 = 7.88

93
i (12)  11.16
100
For P93:

P93 is located at the 12th term P93 = 8.97

 x  61  3.588
N 17
3.10 n = 17; Mean = The mean is 3.588
 P a g e | 80

th
 17  1 
 
 2 
The median is located at the = 9th term, Median = 4

There are eight 4’s, therefore the Mode = 4

75
(17)  12.75
100
Q3 = P75: i=

Q3 is located at the 13th term and Q3 = 4

11
(17)  1.87
100
P11: i=

P11 is located at the 2nd term and P11 = 1

35
(17)  5.95
100
P35: i=

P35 is located at the 6th term and P35 = 3

58
(17)  9.86
100
P58: i=

P58 is located at the 10th term and P58 = 4

 P a g e | 81

67
(17)  11.39
100
P67: i=

P67 is located at the 12th term and P67 = 4

3.11 x   x -µ  (x-µ)2

6 6-4.2857 = 1.7143 2.9388

2 2.2857 5.2244

4 0.2857 0.0816

9 4.7143 22.2246

1 3.2857 10.7958

3 1.2857 1.6530

5 0.7143 0.5102

x = 30  x-µ = 14.2857 (x -µ)2 = 43.4284

x 30
   4.2857
N 7

a.) Range = 9 - 1 = 8

x 14.2857
 
N 7
b.) M.A.D. = 2.0408

( x   ) 2 43.4284

N 7
c.) 2 = = 6.2041
 P a g e | 82

( x   ) 2
 6.2041
N
d.)  = = 2.4908

e.) Arranging the data in order: 1, 2, 3, 4, 5, 6, 9

25
(7)
100
Q1 = P25 i = = 1.75

Q1 is located at the 2nd term, Q1 = 2

75
(7)
100
Q3 = P75: i = = 5.25

Q3 is located at the 6th term, Q3 = 6

IQR = Q3 - Q1 = 6 - 2 = 4

6  4.2857
2.4908
f.) z = = 0.69

2  4.2857
2.4908
z = = -0.92
 P a g e | 83

4  4.2857
2.4908
z = = -0.11

9  4.2857
2.4908
z = = 1.89

1  4.2857
2.4908
z = = -1.32

3  4.2857
2.4908
z = = -0.52

5  4.2857
2.4908
z = = 0.29

xx ( x  x) 2
3.12 x

4 0 0

3 1 1

0 4 16

5 1 1

2 2 4

9 5 25

4 0 0
 P a g e | 84

5 1 1

xx ( x  x ) 2
x = 32 = 14 = 48

x 32
x 
n 8
=4

a) Range = 9 - 0 = 9

xx 14

n 8
b) M.A.D. = = 1.75

( x  x) 2 48

n 1 7
c) s2 = = 6.8571

( x  x) 2
 6.857
n 1
d) s = = 2.6186

e) Numbers in order: 0, 2, 3, 4, 4, 5, 5, 9

25
(8)
100
Q1 = P25 i = = 2

Q1 is located at the average of the 2nd and 3rd terms, Q1

= 2.5
 P a g e | 85

75
(8)
100
Q3 = P75 i = = 6

Q3 is located at the average of the 6th and 7th terms, Q3

=5

IQR = Q3 - Q1 = 5 - 2.5 = 2.5

3.13 a.)

x (x-µ) ( x - µ) 2

12 12-21.167= -9.167 84.034

23 1.833 3.360

19 -2.167 4.696

26 4.833 23.358

24 2.833 8.026

23 1.833 3.360

x = 127 (x -µ) = -0.002 (x -µ)2 = 126.834

x 127

N 6
 = = 21.167

( x   ) 2 126.834
  21.139
N 6
 = = 4.598 ORIGINAL
FORMULA
 P a g e | 86

b.)

x x2

12 144

23 529

19 361

26 676

24 576

23 529

x = 127 x2= 2815

=

(x ) 2 (127) 2
x 2  2815 
N  6 2815  2688.17 126.83
   21.138
N 6 6 6

= 4.598 SHORT-CUT FORMULA

The short-cut formula is faster, but the original

formula gives insight

into the meaning of a standard deviation.

 P a g e | 87

3.14 s2 = 433.9267

s = 20.8309

x = 1387

x2 = 87,365

n = 25

x
= 55.48

3.15  2 = 58,631.295

 = 242.139

x = 6886

x2 = 3,901,664

n = 16

µ = 430.375

3.16

14, 15, 18, 19, 23, 24, 25, 27, 35, 37, 38, 39, 39, 40, 44,

46, 58, 59, 59, 70, 71, 73, 82, 84, 90

 P a g e | 88

25
(25)
100
Q1 = P25 i = = 6.25

P25 is located at the 7th term, and therefore, Q1 = 25

75
(25)
100
Q3 = P75 i = = 18.75

P75 is located at the 19th term, and therefore, Q3 = 59

IQR = Q3 - Q1 = 59 - 25 = 34

1 1 3
1 2
 1    .75
2 4 4
3.17 a) .75

1 1
1 2
 1  .84
2 .5 6.25
b) .84

1 1
1 2
 1  .609
1 .6 2.56
c) .609
 P a g e | 89

1 1
1 2
 1  .902
3 .2 10.24
d) .902
 P a g e | 90

3.18

Set 1:

x 262
1    65.5
N 4

(x ) 2 (262) 2
x 2  17,970 
1  N  4
N 4
= 14.2215

Set 2:

x 570
2    142.5
N 4

(x ) 2 (570) 2
x 2  82,070 
2  N  4
N 4
= 14.5344

14.2215
(100)
65.5
CV1 = = 21.71%
 P a g e | 91

14.5344
(100)
142.5
CV2 = = 10.20%
 P a g e | 92

x xx ( x  x) 2
3.19

7 1.833 3.361

5 3.833 14.694

10 1.167 1.361

12 3.167 10.028

9 0.167 0.028

8 0.833 0.694

14 5.167 26.694

3 5.833 34.028

11 2.167 4.694

13 4.167 17.361

8 0.833 0.694

6 2.833 8.028

106 32.000
121.665

x 106
x 
n 12
= 8.833

xx 32

n 12
a) MAD = = 2.667

( x  x) 2 121.665

n 1 11
b) s2 = = 11.06
 P a g e | 93

s 2  11.06
c) s = = 3.326

d) Rearranging terms in order: 3 5 6 7 8 8 9 10 11 12 13 14

Q1 = P25: i = (.25)(12) = 3

Q1 = the average of the 3rd and 4th terms: Q1 = (6 + 7)/2 = 6.5

Q3 = P75: i = (.75)(12) = 9

Q3 = the average of the 9th and 10th terms: Q3 = (11 + 12)/2 =

11.5

IQR = Q3 - Q1 = 11.5 – 6.5 = 5

6  8.833
3.326
e.) z = = - 0.85

(3.326)(100)
8.833
f.) CV = = 37.65%

3.20 n = 11 x   x- 

768 475.64
429 136.64

323 30.64

306 13.64
 P a g e | 94

286 6.36

262 30.36

215 77.36

172 120.36

162 130.36
148 144.36

145 147.36

x = 3216  x-µ = 1313.08

µ = 292.36 x = 3216 x2 = 1,267,252

a.) Range = 768 - 145 = 623

x 1313.08

N 11
b.) MAD = = 119.37

(x ) 2 (3216) 2
x  2
1,267,252 
N  11
N 11
c.) 2 = = 29,728.23

29,728.23
d.)  = = 172.42

e.) Q1 = P25: i = .25(11) = 2.75

Q1 is located at the 3rd term and Q1 = 162

 P a g e | 95

Q3 = P75: i = .75(11) = 8.25

Q3 is located at the 9th term and Q3 = 323

IQR = Q3 - Q1 = 323 - 162 = 161

f.) xnestle = 172

x   172  292.36

 172.42
z = = -0.70

 172.42
(100)  (100)
 292.36
g.) CV = = 58.98%

3.21 µ = 125  = 12

68% of the values fall within:

µ ± 1 = 125 ± 1(12) = 125 ± 12

between 113 and 137

95% of the values fall within:

µ ± 2 = 125 ± 2(12) = 125 ± 24

between 101 and 149

 P a g e | 96

99.7% of the values fall within:

µ ± 3 = 125 ± 3(12) = 125 ± 36

between 89 and 161

3.22 µ = 38 =6

between 26 and 50:

x1 - µ = 50 - 38 = 12

x2 - µ = 26 - 38 = -12

x1   12

 6
= 2

x 2    12

 6
= -2

k = 2, and since the distribution is not normal, use Chebyshev’s

theorem:

1 1 1 3
1 2
 1 2  1 
k 2 4 4
= .75
 P a g e | 97

at least 75% of the values will fall between 26 and 50

 P a g e | 98

between 14 and 62? µ = 38  =6

x1 - µ = 62 - 38 = 24

x2 - µ = 14 - 38 = -24

x1   24

 6
= 4

x 2    24

 6
= -4

k=4

1 1 1 15
1 2
 1 2  1 
k 4 16 16
= .9375

at least 93.75% of the values fall between 14 and 62

between what 2 values do at least 89% of the values fall?

1
k2
1- =.89
 P a g e | 99

1
k2
.11 =

.11 k2 = 1

1
.11
k2 =

k2 = 9.09

k = 3.015

With µ = 38,  = 6 and k = 3.015 at least 89% of the values fall

within:

µ ± 3.015 = 38 ± 3.015 (6) = 38 ± 18.09

Between 19.91 and 56.09

1
k2
3.23 1- = .80

1
k2
1 - .80 =
 P a g e | 100

1
k2
.20 = and .20k2 = 1

k2 = 5 and k = 2.236

2.236 standard deviations

3.24 µ = 43. 68% of the values lie µ + 1 . Thus, between the mean, 43,
and one of

the values, 46, is one standard deviation. Therefore,

1 = 46 - 43 = 3

within 99.7% of the values lie µ + 3. Thus, between the mean, 43,
and one of

the values, 51, are three standard deviations. Therefore,

3 = 51 - 43 = 8

 = 2.67

µ = 28 and 77% of the values lie between 24 and 32 or + 4 from the

mean:
 P a g e | 101

1
k2
1- = .77

Solving for k:

1
k2
.23 = and therefore, .23k2 = 1

k2 = 4.3478

k = 2.085

2.085 = 4

4
2.085
 = = 1.918

3.25 µ = 29 =4

Between 21 and 37 days:

x1   21  29  8
 
 4 4
= -2 Standard Deviations

x 2   37  29 8
 
 8 4
= 2 Standard Deviations
 P a g e | 102

Since the distribution is normal, the empirical rule states that

95% of

the values fall within µ ± 2 .

Exceed 37 days:

Since 95% fall between 21 and 37 days, 5% fall outside this range.
Since the

normal distribution is symmetrical, 2½% fall below 21 and above 37.

Thus, 2½% lie above the value of 37.

Exceed 41 days:

x   41  29 12
 
 4 4
= 3 Standard deviations

The empirical rule states that 99.7% of the values fall within µ ± 3 =
29 ± 3(4) =

29 ± 12. That is, 99.7% of the values will fall between 17 and 41 days.

0.3% will fall outside this range and half of this or .15% will lie
above 41.
 P a g e | 103

Less than 25: µ = 29 = 4

x   25  29  4
 
 4 4
= -1 Standard Deviation

According to the empirical rule, µ ± 1 contains 68% of the values.

29 ± 1(4) = 29 ± 4

Therefore, between 25 and 33 days, 68% of the values lie and 32% lie
outside this

range with ½(32%) = 16% less than 25.

3.26 x

97

109

111

118

120

130

132

133

137

137

x
x = 1224 x = 151,486
2
n = 10 = 122.4 s = 13.615
 P a g e | 104

Bordeaux: x = 137

137  122.4
13.615
z = = 1.07

Montreal: x = 130

130  122.4
13.615
z = = 0.56

Edmonton: x = 111

111  122.4
13.615
z = = -0.84

Hamilton: x = 97

97  122.4
13.615
z = = -1.87
 P a g e | 105

3.27 Mean

Class f M fM

0- 2 39 1 39

2- 4 27 3 81

4- 6 16 5 80

6- 8 15 7 105

8 - 10 10 9 90

10 - 12 8 11 88

12 - 14 6 13 78

f=121 fM=561

fM 561

f 121
 = = 4.64

Mode: The modal class is 0 – 2.

The midpoint of the modal class = the mode = 1

3.28

Class f M fM

1.2 - 1.6 220 1.4 308

1.6 - 2.0 150 1.8 270

2.0 - 2.4 90 2.2 198

2.4 - 2.8 110 2.6 286

 P a g e | 106

2.8 - 3.2 280 3.0 840

f=850 fM=1902

fM 1902

f 850
Mean:  = = 2.24

Mode: The modal class is 2.8 – 3.2.

The midpoint of the modal class is the mode = 3.0

 P a g e | 107

3.29 Class f M fM

20-30 7 25 175

30-40 11 35 385

40-50 18 45 810

50-60 13 55 715

60-70 6 65 390

70-80 4 75 300

Total 59 2775

fM 2775

f 59
 = = 47.034

M-µ ( M - µ) 2 f ( M - µ) 2

-22.0339 485.4927 3398.449

-12.0339 144.8147 1592.962

- 2.0339 4.1367 74.462

7.9661 63.4588 824.964

17.9661 322.7808 1936.685

27.9661 782.1028 3128.411

Total 10,955.933

f ( M   ) 2 10,955.93

f 59
2 = = 185.694

185.694
 = = 13.627
 P a g e | 108

3.30 Class f M fM fM2

5- 9 20 7 140 980

9 - 13 18 11 198 2,178

13 - 17 8 15 120 1,800

17 - 21 6 19 114 2,166

21 - 25 2 23 46 1,058

f=54 fM= 618 fM2= 8,182

(fM ) 2 (618) 2
fM 2  8182 
n  54  8182  7071.67
n 1 53 53
s2 = = 20.931

s 2  20.9
s = = 4.575
 P a g e | 109

3.31 Class f M fM fM2

18 - 24 17 21 357
7,497

24 - 30 22 27 594
16,038

30 - 36 26 33 858
28,314

36 - 42 35 39 1,365
53,235

42 - 48 33 45 1,485
66,825

48 - 54 30 51 1,530
78,030

54 - 60 32 57 1,824
103,968

60 - 66 21 63 1,323
83,349

66 - 72 15 69 1,035
71,415

f= 231 fM= 10,371 fM2= 508,671

fM fM 10,371

x  
n f 231
a.) Mean: = 44.896

b.) Mode. The Modal Class = 36-42. The mode is the class midpoint =
39
 P a g e | 110

(fM ) 2 (10,371) 2
fM 
2
508,671 
n  231  43,053.5065
n 1 230 230
c.) s2 = = 187.189

187.2
d.) s = = 13.682
 P a g e | 111

3.32
Class f M fM fM2

0-1 31 0.5 15.5 7.75

1-2 57 1.5 85.5 128.25

2-3 26 2.5 65.0 162.50

3-4 14 3.5 49.0 171.50

4-5 6 4.5 27.0 121.50

5-6 3 5.5 16.5 90.75

f=137 fM=258.5 fM2=682.25

a.) Mean

fM 258.5

f 137
 = = 1.887

b.) Mode: Modal Class = 1-2. Mode = 1.5

c.) Variance:

(fM ) 2 ( 258.5) 2
fM 2  682.25 
N  137
N 137
2 = = 1.4197

d.) Standard Deviation:

 2  1.4197
 = = 1.1915
P a g e | 112
 P a g e | 113

3.33 f M fM fM2

20-30 8 25 200 5000

30-40 7 35 245 8575

40-50 1 45 45 2025

50-60 0 55 0 0

60-70 3 65 195 12675

70-80 1 75 75 5625

f = 20 fM = 760 fM2 = 33900

a.) Mean:

fM 760

f 20
 = = 38

b.) Mode. The Modal Class = 20-30.

The mode is the midpoint of this class = 25.

c.) Variance:

(fM ) 2 (760) 2
fM 2  33,900 
N  20
N 20
2 = = 251

d.) Standard Deviation:

 P a g e | 114

 2  251
 = = 15.843
 P a g e | 115

3.34 No. of Farms f M fM

0 - 20,000 16 10,000 160,000

20,000 - 40,000 11 30,000 330,000

40,000 - 60,000 10 50,000 500,000

60,000 - 80,000 6 70,000 420,000

80,000 - 100,000 5 90,000 450,000

100,000 - 120,000 1 110,000 110,000

f = 49 fM = 1,970,000

fM 1,970,000

f 49
 = = 40,204

The actual mean for the ungrouped data is 37,816. This computed
group

mean, 40,204, is really just an approximation based on using the class

midpoints in the calculation. Apparently, the actual numbers of farms

per

state in some categories do not average to the class midpoint and in

fact

might be less than the class midpoint since the actual mean is less
than the

grouped data mean.

The value for fM2 is 1.185x1011

(fm) 2 (1,970,000) 2
fM 
2
1.185 x10 
11

N  49
N 49
2 = = 801,999,167
 P a g e | 116

 = 28,319.59

The actual standard deviation was 29,341. The difference again is due
to

the grouping of the data and the use of class midpoints to represent
the

data. The class midpoints due not accurately reflect the raw data.

3.35 mean = $35

median = $33

mode = $21

The stock prices are skewed to the right. While many of the stock
prices

are at the cheaper end, a few extreme prices at the higher end pull the

mean.
 P a g e | 117

3.36 mean = 51

median = 54

mode = 59

The distribution is skewed to the left. More people are older but the
most

extreme ages are younger ages.

3(   M d ) 3(5.51  3.19)

 9.59
3.37 Sk = = 0.726

3.38 n = 25 x = 600

x
= 24 s = 6.6521 Md = 23

3( x  M d ) 3(24  23)

s 6.6521
Sk = = 0.451

There is a slight skewness to the right

3.39 Q1 = 500. Median = 558.5. Q3 = 589.

 P a g e | 118

IQR = 589 - 500 = 89

Inner Fences: Q1 - 1.5 IQR = 500 - 1.5 (89) = 366.5

and Q3 + 1.5 IQR = 589 + 1.5 (89) = 722.5

Outer Fences: Q1 - 3.0 IQR = 500 - 3 (89) = 233

and Q3 + 3.0 IQR = 589 + 3 (89) = 856

The distribution is negatively skewed. There are no mild or extreme

outliers.
 P a g e | 119

3.40 n = 18
th
(n  1)th (18  1)th 19
 
2 2 2
Median: = 9.5th term

Median = 74

Q1 = P25:

25
(18)
100
i = = 4.5

Q1 = 5th term = 66

Q3 = P75:

75
(18)
100
i = = 13.5

Q3 = 14th term = 90

Therefore, IQR = Q3 - Q1 = 90 - 66 = 24

Inner Fences: Q1 - 1.5 IQR = 66 - 1.5 (24) = 30

Q3 + 1.5 IQR = 90 + 1.5 (24) = 126

 P a g e | 120

Outer Fences: Q1 - 3.0 IQR = 66 - 3.0 (24) = -6

Q3 + 3.0 IQR = 90 + 3.0 (24) = 162

There are no extreme outliers. The only mild outlier is 21. The

distribution is positively skewed since the median is nearer to Q1 than

Q3.
 P a g e | 121

3.41 x = 80 x2 = 1,148 y = 69

y2 = 815 xy = 624 n=7

 x y
 xy  n
 ( x )   2
( y ) 2 
 x  y 
2 2
  
 n   n 
r = =

(80)(69)
624 
7
 (80)  
2
(69) 2   164.571
 1,148    815  
 7   7  ( 233.714)(134.857)
r = = =

 164.571
177.533
r = = -0.927

3.42 x = 1,087 x2 = 322,345 y = 2,032

y2 = 878,686 xy= 507,509 n=5

 P a g e | 122

 x y
 xy  n
 ( x ) 2   ( y ) 2 
 x   y2 
2
  
 n   n 
r = =

(1,087)( 2,032)
507,509 
5
 (1,087)  
2
(2,032) 2 
 322,345    878, 686  
 5   5 
r = =

65,752.2 65,752.2
(86,031.2)(52,881.2) 67,449.5
r = = = .975
 P a g e | 123

3.43 Delta (x) SW (y)

47.6 15.1

46.3 15.4

50.6 15.9
52.6 15.6

52.4 16.4

52.7 18.1

x = 302.2 y = 96.5 xy = 4,870.11

x2 = 15,259.62 y2 = 1,557.91

 x y
 xy  n
 ( x ) 2   ( y ) 2 
 x   y2 
2
  
 n   n 
r = =

(302.2)(96.5)
4,870.11 
6
 (302.2) 2   (96.5) 2 
 15, 259. 62    1,557 .91  
 6   6 
r = = .6445

3.44 x = 6,087 x2 = 6,796,149

y = 1,050 y2 = 194,526

xy = 1,130,483 n=9

 P a g e | 124

 x y
 xy  n
 ( x )   2
( y ) 2 
 x  y 
2 2
  
 n   n 
r = =

(6,087)(1,050)
1,130,483 
9
 (6,087)  
2
(1,050) 2 
 6, 796,149    194 ,526  
 9   9 
r = =

420,333 420,333
(2,679,308)(72,026) 439,294.705
r = = = .957

3.45 Correlation between Year 1 and Year 2:

x = 17.09 x2 = 58.7911

y = 15.12 y2 = 41.7054

xy = 48.97 n=8

 x y
 xy  n
 ( x )   2
( y ) 2 
 x  y 
2 2
  
 n   n 
r = =
 P a g e | 125

(17.09)(15.12)
48.97 
8
 (17.09)  
2
(15.12) 2 
 58. 7911    41 . 7054  
 8   8 
r = =

16.6699 16.6699
(22.28259)(13.1286) 17.1038
r = = = .975

Correlation between Year 2 and Year 3:

x = 15.12 x2 = 41.7054

y = 15.86 y2 = 42.0396

xy = 41.5934 n=8

 x y
 xy  n
 ( x ) 2   ( y ) 2 
 x   y2 
2
  
 n   n 
r = =

(15.12)(15.86)
41.5934 
8
 (15.12)  
2
(15.86) 2 
 41. 7054    42 . 0396  
 8   8 
r = =
 P a g e | 126

11.618 11.618
(13.1286)(10.59715) 11.795
r = = = .985

Correlation between Year 1 and Year 3:

x = 17.09 x2 = 58.7911

y = 15.86 y2 = 42.0396

xy = 48.5827 n=8

 x y
 xy  n
 ( x ) 2   ( y ) 2 
 x   y2 
2
  
 n   n 
r = =

(17.09)(15.86)
48.5827 
8
 (17.09) 2   (15.86) 2 
 58.7911    42.0396  
 8   8 
r =

14.702 14.702
( 22.2826)(10.5972) 15.367
r = = = .957
 P a g e | 127

The years 2 and 3 are the most correlated with r = .985.

 P a g e | 128

3.46 Arranging the values in an ordered array:

1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,

3, 3, 3, 3, 3, 3, 4, 4, 5, 6, 8

x 75
x 
n 30
Mean: = 2.5

Mode = 2 (There are eleven 2’s)

Median: There are n = 30 terms.

th
n 1 30  1 31
 
2 2 2
The median is located at = 15.5th position.

Median is the average of the 15th and 16th value.

However, since these are both 2, the median is 2.

Range = 8 - 1 = 7

25
(30)
100
Q1 = P25: i = = 7.5
 P a g e | 129

Q1 is the 8th term = 1

75
(30)
100
Q3 = P75: i = = 22.5

Q3 is the 23rd term = 3

IQR = Q3 - Q1 = 3 - 1 = 2
 P a g e | 130

10
( 40)
100
3.47 P10: i = = 4

P10 = 4.5th term = 23

80
( 40)
100
P80: i = = 32

P80 = 32.5th term = 49.5

25
( 40)
100
Q1 = P25: i = = 10

P25 = 10.5th term = 27.5

75
( 40)
100
Q3 = P75: i = = 30

P75 = 30.5th term = 47.5

IQR = Q3 - Q1 = 47.5 - 27.5 = 20

 P a g e | 131

Range = 81 - 19 = 62
 P a g e | 132

x 126,904

N 20
3.48 µ = = 6345.2

n 1
2
The median is located at the th value = 21/2 = 10.5th value

The median is the average of 5414 and 5563 = 5488.5

P30: i = (.30)(20) = 6

P30 is located at the average of the 6th and 7th terms

P30 = (4507+4541)/2 = 4524

P60: i = (.60)(20) = 12

P60 is located at the average of the 12th and 13th terms

P60 = (6101+6498)/2 = 6299.5

P90: i = (.90)(20) = 18

P90 is located at the average of the 18th and 19th terms

P90 = (9863+11,019)/2 = 10,441

 P a g e | 133

Q1 = P25: i = (.25)(20) = 5

Q1 is located at the average of the 5th and 6th terms

Q1 = (4464+4507)/2 = 4485.5

Q3 = P75: i = (.75)(20) = 15

Q3 is located at the average of the 15th and 16th terms

Q3 = (6796+8687)/2 = 7741.5

Range = 11,388 - 3619 = 7769

IQR = Q3 - Q1 = 7741.5 - 4485.5 = 3256

3.49 n = 10 x = 87.95 x2 = 1130.9027

µ = (x)/N = 87.95/10 = 8.795

( x ) 2 (87.95) 2
x 2  1130.9027 
N  10
N 10
 = = 5.978
 P a g e | 134

x
N
3.50 a.) µ= = 26,675/11 = 2425

Median = 1965

b.) Range = 6300 - 1092 = 5208

Q3 = 2867 Q1 = 1532 IQR = Q3 - Q1 = 1335

c.) Variance:

(x ) 2 (26,675) 2
x 2  86,942,873 
N  11
N 11
2 = = 2,023,272.55

Standard Deviation:

 2  2,023,272.55
 = = 1422.42

d.) Texaco:

x   1532  2425

 1422.42
z = = -0.63
 P a g e | 135

Exxon Mobil:

x   6300  2425

 1422.42
z = = 2.72

e.) Skewness:

3(   M d ) 3(2425  1965)

 1422.42
Sk = = 0.97

x 32.95
 
n 14
3.51 a.) Mean: = 2.3536

1.79  2.07
2
Median: = 1.93

Mode: No Mode

b.) Range: 4.73 – 1.20 = 3.53

1
(14)  3.5
4
Q1: Located at the 4th term. Q1 = 1.68

3
(14)  10.5
4
Q3: Located at the 11th term. Q3 = 2.87
 P a g e | 136

IQR = Q3 – Q1 = 2.87 – 1.68 = 1.19

xx  x  x 2

4.73 2.3764 5.6473

3.64 1.2864 1.6548

3.53 1.1764 1.3839
2.87 0.5164 0.2667
2.61 0.2564 0.0657
2.59 0.2364 0.0559
2.07 0.2836 0.0804
1.79 0.5636 0.3176
1.77 0.5836 0.3406
1.69 0.6636 0.4404
1.68 0.6736 0.4537
1.41 0.9436 0.8904
1.37 0.9836 0.9675
1.20 1.1536 1.3308
 xx  ( x  x) 2

= 11.6972 = 13.8957

 xx 
11.6972
n 14
MAD = = 0.8355

  x  x
2
13.8957

n 1 13
s2 = = 1.0689

s 2  1.0689
s= = 1.0339

c.) Pearson’s Coefficient of Skewness:

3( x  M d ) 3(2.3536  1.93)

s 1.0339
Sk = = 1.229

d.) Use Q1 = 1.68, Q2 = 1.93, Q3 = 2.87, IQR = 1.19

Extreme Points: 1.20 and 4.73

 P a g e | 137

Inner Fences: 1.68 – 1.5(1.19) = -0.105

1.93 + 1.5(1.19) = 3.715

Outer Fences: 1.68 + 3.0(1.19) = -1.890

1.93 + 3.0(1.19) = 5.500

There is one mild outlier. The 4.73 recorded for Arizona is outside
the upper
inner fence.

Mineral Production

1 2 3 4 5
Mineral Production
 P a g e | 138

3.52 f M fM
2
fM

15-20 9 17.5 157.5 2756.25

20-25 16 22.5 360.0 8100.00

25-30 27 27.5 742.5 20418.75

30-35 44 32.5 1430.0 46475.00

35-40 42 37.5 1575.0 59062.50

40-45 23 42.5 977.5 41543.75

45-50 7 47.5 332.5 15793.75

50-55 2 52.5 105.0 5512.50

f = 170 fM = 5680.0 fM2 = 199662.50

a.) Mean:

fM 5680

f 170
 = = 33.412

Mode: The Modal Class is 30-35. The class midpoint is the mode =
32.5.

b.) Variance:
 P a g e | 139

(fM ) 2 (5680) 2
fM  2
199,662.5 
n  170
n 1 169
s2 = = 58.483

Standard Deviation:

s 2  58.483
s = = 7.647
 P a g e | 140

3.53 Class f M fM fM2

0 - 20 32 10 320 3,200

20 - 40 16 30 480 14,400

40 - 60 13 50 650 32,500

60 - 80 10 70 700 49,000

80 - 100 19 90 1,710 153,900

f = 90 fM= 3,860 fM2= 253,000

fM fm 3,860

x  
n f 90
a) Mean: = 42.89

Mode: The Modal Class is 0-20. The midpoint of this class is the
mode = 10.

b) Sample Standard Deviation:

(fM ) 2 (3860) 2
fM 2  253,000 
n 90 253,000  165,551.1
 
n 1 89 89
87,448.9
  982.572
89
s =

= 31.346
 P a g e | 141

3.54 x = 36 x2 = 256

y = 44 y2 = 300

xy = 188 n =7

 x y
 xy  n 188 
(36)( 44)
7
 ( x )  
2
( y ) 2   (36)  
2
(44) 
 x   y2 
2
    256    300  
 n   n   7   7 
r = =

 38.2857  38.2857
(70.85714)( 23.42857) 40.7441
r = = = -.940
 P a g e | 142

x 3.45
(100%)  (100%)
x 32
3.55 CVx = = 10.78%

y 5.40
(100%)  (100%)
y 84
CVY = = 6.43%

Stock X has a greater relative variability.

3.56 µ = 7.5 Each of the numbers, 1 and 14, are 6.5 units away from
the mean.

From the Empirical Rule: 99.7% of the values lie in µ + 3

3 = 14 - 7.5 = 6.5 Solving for 3 = 6.5 for :  = 2.167

Suppose that µ = 7.5,  = 1.7:

95% lie within µ + 2 = 7.5 + 2(1.7) = 7.5 + 3.4

Between 4.1 and 10.9 lie 95% of the values.

 P a g e | 143

3.57 µ = 419,  = 27

a.) 68%: µ + 1 419 + 27 392 to 446

95%: µ + 2 419 + 2(27) 365 to 473

99.7%: µ + 3 419 + 3(27) 338 to 500

b.) Use Chebyshev’s:

Each of the points, 359 and 479 is a distance of 60 from the mean,
µ = 419.

k = (distance from the mean)/ = 60/27 = 2.22

Proportion = 1 - 1/k2 = 1 - 1/(2.22)2 = .797 = 79.7%

400  419
27
c.) Since x = 400, z = = -0.704. This worker is in the lower
half of

workers but within one standard deviation of the mean.

3.58 a.) x x2

Albania 4,900 24,010,000

Bulgaria 8,200 67,240,000

Croatia 11,200 125,440,000

Czech 16,800 282,240,000

 P a g e | 144

x=41,100 x2= 498,930,000

x 41,100

N 4
 = = 10,275

(x ) 2 (41,100) 2
x 2  498,930,000 
N  4
N 4
 = = 4376.86

b.) x x2

Hungary 14,900 222,010,000

Poland 12,000 144,000,000

Romania 7,700 59,290,000

Bosnia/Herz 6,500 42,250,000

x=41,100 x2=467,550,000

x 41,100

N 4
 = = 10,275

(x ) 2 (41,100) 2
x 2  467,550,000 
N  4
N 4
 = = 3363.31

c.)
 P a g e | 145

1 4376.86
(100)  (100)
1 10,275
CV1 = = 42.60%

2 3363.31
(100)  (100)
2 10,275
CV2 = = 32.73%

The first group has a larger coefficient of variation

 P a g e | 146

3.59 Mean $35,748

Median $31,369

Mode $29,500

Since these three measures are not equal, the distribution is skewed.
The

distribution is skewed to the right because the mean is greater than

the median. Often, the median is preferred in reporting income data
because it yields information about the middle of the data while
ignoring extremes.

3.60 x = 36.62 x2 = 217.137

y = 57.23 y2 = 479.3231

xy = 314.9091 n =8

 x y
 xy  n
 ( x )   2
( y ) 2 
 x  y 
2 2
  
 n   n 
r = =

(36.62)(57.23)
314.9091 
8
 (36.62)  
2
(57.23) 2 
 217.137    479.3231  
 8   8 
r = =
 P a g e | 147

52.938775
(49.50895)(69.91399)
r = = .90

There is a strong positive relationship between the inflation rate and

the thirty-year treasury yield.

 P a g e | 148

25
(20)
100
3.61 a.) Q1 = P25: i = = 5

Q1 = 5.5th term = (48.3 + 49.9)/2 = 49.1

75
(20)
100
Q3 = P75: i = = 15

Q3 = 15.5th term = (77.6+83.8)/2 = 80.7

th th
n 1 20  1

2 2
Median: = 10.5th term

Median = (55.9 + 61.3)/2 = 58.6

IQR = Q3 - Q1 = 80.7 – 49.1 = 31.6

1.5 IQR = 47.4; 3.0 IQR = 94.8

Inner Fences:

Q1 - 1.5 IQR = 49.1 – 47.4 = 1.7

Q3 + 1.5 IQR = 80.7 + 47.4 = 128.1

Outer Fences:
 P a g e | 149

Q1 - 3.0 IQR = 49.1 – 94.8 = - 45.70

Q3 + 3.0 IQR = 80.7 + 94.8 = 175.5

b.) and c.) There are no outliers in the lower end. There are two
extreme

outliers in the upper end (South Louisiana, 198.8, and

Houston,

190.9). There is one mild outlier at the upper end (New York,
145.9).

Since the median is nearer to Q1, the distribution is

positively skewed.

d.) There are three dominating, large

ports

Displayed below is the MINITAB boxplot for this problem.

 P a g e | 150

Boxplot of U.S. Ports

50 100 150 200

U.S. Ports

3.62 Paris: Since 1 - 1/k2 = .53, solving for k: k = 1.459

The distance from µ = 349 to x = 381 is 32

1.459 = 32

 = 21.93

Moscow: Since 1 - 1/k2 = .83, solving for k: k = 2.425

The distance from µ = 415 to x = 459 is 44

 P a g e | 151

2.425 = 44

 = 18.14

Chapter 4
Probability

LEARNING OBJECTIVES

The main objective of Chapter 4 is to help you understand the basic principles of
probability, specifically enabling you to

1. Comprehend the different ways of assigning probability.

2. Understand and apply marginal, union, joint, and conditional

probabilities.

3. Select the appropriate law of probability to use in solving problems.

 P a g e | 152

4. Solve problems using the laws of probability, including the law of

addition, the law of multiplication , and the law of conditional
probability.

5. Revise probabilities using Bayes' rule.

CHAPTER TEACHING STRATEGY

Students can be motivated to study probability by realizing that the

field of probability has some stand-alone application in their lives in such
applied areas human resource analysis, actuarial science, and gaming. In
addition, students should understand that much of the rest of this course is
based on probabilities even though they will not be directly applying many of
these formulas in other chapters.

This chapter is frustrating for the learner because probability problems

can be approached by using several different techniques. Whereas, in many
chapters of this text, students will approach problems by using one standard
technique, in chapter 4, different students will often use different approaches
to the same problem. The text attempts to emphasize this point and
underscore it by presenting several different ways to solve probability
problems. The probability rules and laws presented in the chapter can
virtually always be used in solving probability problems. However, it is
sometimes easier to construct a probability matrix or a tree diagram or use
the sample space to solve the problem. If the student is aware that what
they have at their hands is an array of tools or techniques, they will be less
overwhelmed in approaching a probability problem. An attempt has been
made to differentiate the several types of probabilities so that students can
sort out the various types of problems.

In teaching students how to construct a probability matrix, emphasize

that it is usually best to place only one variable along each of the two
dimensions of the matrix. (That is, place Mastercard with yes/no on one axis
and Visa with yes/no on the other instead of trying to place Mastercard and
Visa along the same axis).

This particular chapter is very amenable to the use of visual aids.

Students enjoy rolling dice, tossing coins, and drawing cards as a part of the
class experience. Of all the chapters in the book, it is most imperative that
students work a lot of problems in this chapter. Probability problems are so
varied and individualized that a significant portion of the learning comes in
the doing. Experience is an important factor in working probability problems.
 P a g e | 153

Section 4.8 on Bayes’ theorem can be skipped in a one-semester

course without losing any continuity. This section is a prerequisite to the
chapter 19 presentation of “revising probabilities in light of sample
information (section 19.4).

CHAPTER OUTLINE

4.1 Introduction to Probability

4.2 Methods of Assigning Probabilities

Classical Method of Assigning Probabilities

Relative Frequency of Occurrence

Subjective Probability

4.3 Structure of Probability

Experiment

Event

Elementary Events

Sample Space

Unions and Intersections

Mutually Exclusive Events

Independent Events

Collectively Exhaustive Events

Complimentary Events

Counting the Possibilities

The mn Counting Rule

Sampling from a Population with Replacement

Combinations: Sampling from a Population Without

Replacement
 P a g e | 154

4.4 Marginal, Union, Joint, and Conditional Probabilities

4.5 Addition Laws

Probability Matrices

Complement of a Union

Special Law of Addition

4.6 Multiplication Laws

General Law of Multiplication

Special Law of Multiplication

4.7 Conditional Probability

Independent Events

4.8 Revision of Probabilities: Bayes' Rule

KEY TERMS

A Priori Intersection

Bayes' Rule Joint Probability

Classical Method of Assigning Probabilities Marginal Probability

Collectively Exhaustive Events mn Counting Rule

Combinations Mutually Exclusive Events

Complement of a Union Probability Matrix

 P a g e | 155

Complement Relative Frequency of

Occurrence

Conditional Probability Sample Space

Elementary Events Set Notation

Event Subjective Probability

Experiment Union

Independent Events Union Probability

 P a g e | 156

SOLUTIONS TO PROBLEMS IN CHAPTER 4

4.1 Enumeration of the six parts: D1, D2, D3, A4, A5, A6

D = Defective part

A = Acceptable part

Sample Space:

D1 D2, D2 D3, D3 A5

D1 D3, D2 A4, D3 A6

D1 A4, D2 A5, A4 A5

D1 A5, D2 A6, A4 A6

D1 A6, D3 A4, A5 A6

There are 15 members of the sample space

The probability of selecting exactly one defect out of

two is:

9/15 = .60

4.2 X = {1, 3, 5, 7, 8, 9}, Y = {2, 4, 7, 9}, and Z = {1, 2, 3, 4, 7,}

 P a g e | 157

a) X  Z = {1, 2, 3, 4, 5, 7, 8, 9}

b) X  Y = {7, 9}

c) X  Z = {1, 3, 7}

d) X  Y  Z = {1, 2, 3, 4, 5, 7, 8, 9}

e) X Y Z= {7}

f) (X  Y)  Z = {1, 2, 3, 4, 5, 7, 8, 9}  {1, 2, 3, 4, 7} = {1, 2,

3, 4, 7}

g) (Y  Z)  (X  Y) = {2, 4, 7}  {7, 9} = {2, 4, 7, 9}

h) X or Y = X  Y = {1, 2, 3, 4, 5, 7, 8, 9}

i) Y and Z = Y  Z = {2, 4, 7}

4.3 If A = {2, 6, 12, 24} and the population is the positive even numbers
through 30,

A’ = {4, 8, 10, 14, 16, 18, 20, 22, 26, 28, 30}

4.4 6(4)(3)(3) = 216

4.5 Enumeration of the six parts: D1, D2, A1, A2, A3, A4

D = Defective part

A = Acceptable part

Sample Space:
 P a g e | 158

D1 D2 A1, D1 D2 A2, D1 D2 A3,

D1 D2 A4, D1 A1 A2, D1 A1 A3,

D1 A1 A4, D1 A2 A3, D1 A2 A4,

D1 A3 A4, D2 A1 A2, D2 A1 A3,

D2 A1 A4, D2 A2 A3, D2 A2 A4,

D2 A3 A4, A1 A2 A3, A1 A2 A4,

A1 A3 A4, A2 A3 A4

Combinations are used to counting the sample space because

sampling is done without replacement.

6!
3!3!
6 C3 = = 20

Probability that one of three is defective is:

12/20 = 3/5 .60

There are 20 members of the sample space and 12 of them have

exactly

1 defective part.

4.6 107 = 10,000,000 different numbers

 P a g e | 159

20!
6!14!
4.7 C6 =
20 = 38,760

It is assumed here that 6 different (without replacement) employees

are to be selected.

4.8 P(A) = .10, P(B) = .12, P(C) = .21, P(A  C) = .05 P(B  C) = .03

a) P(A  C) = P(A) + P(C) - P(A  C) = .10 + .21 - .05 = .26

b) P(B  C) = P(B) + P(C) - P(B  C) = .12 + .21 - .03 = .30

c) If A, B mutually exclusive, P(A  B) = P(A) + P(B) = .10 + .12 = .22

4.9

D E F

A 5 8 12 25

B 10 6 4 20

C 8 2 5 15

23 16 21 60

a) P(A  D) = P(A) + P(D) - P(A  D) = 25/60 + 23/60 - 5/60 =

43/60 = .7167

b) P(E  B) = P(E) + P(B) - P(E  B) = 16/60 + 20/60 - 6/60 = 30/60

= .5000

c) P(D  E) = P(D) + P(E) = 23/60 + 16/60 = 39/60 = .6500

 P a g e | 160

d) P(C  F) = P(C) + P(F) - P(C  F) = 15/60 + 21/60 - 5/60 = 31/60

= .5167
 P a g e | 161

4.10

E F

A .10 .03 .13

B .04 .12 .16

C .27 .06 .33

D .31 .07 .38

.72 .28 1.00

a) P(A  F) = P(A) + P(F) - P(A  F) = .13 + .28 - .03 = .38

b) P(E  B) = P(E) + P(B) - P(E  B) = .72 + .16 - .04 = .84

c) P(B  C) = P(B) + P(C) =.16 + .33 = .49

d) P(E  F) = P(E) + P(F) = .72 + .28 = 1.000

4.11 A = event of having flown in an airplane at least once

T = event of having ridden in a train at least once

P(A) = .47 P(T) = .28

P (ridden either a train or an airplane) =

P(A  T) = P(A) + P(T) - P(A  T) = .47 + .28 - P(A  T)

Cannot solve this problem without knowing the probability of the

intersection.
 P a g e | 162

We need to know the probability of the intersection of A and T, the

proportion

who have ridden both or determine if these two events are mutually exclusive.

4.12 P(L) = .75 P(M) = .78 P(M  L) = .61

a) P(M  L) = P(M) + P(L) - P(M  L) = .78 + .75 - .61 = .92

b) P(M  L) but not both = P(M  L) - P(M  L) = .92 - .61 = .31

c) P(NM  NL) = 1 - P(M  L) = 1 - .92 = .08

Note: the neither/nor event is solved for here by taking the

complement of the

union.
 P a g e | 163

4.13 Let C = have cable TV

Let T = have 2 or more TV sets

P(C) = .67, P(T) = .74, P(C  T) = .55

a) P(C  T) = P(C) + P(T) - P(C  T) = .67 + .74 - .55 = .86

b) P(C  T but not both) = P(C  T) - P(C  T) = .86 - .55 = .31
c) P(NC  NT) = 1 - P(C  T) = 1 - .86 = .14
d) The special law of addition does not apply because P(C  T) is not .0000.
Possession of cable TV and 2 or more TV sets are not mutually exclusive.

4.14 Let T = review transcript

F = consider faculty references

P(T) = .54 P(F) = .44 P(T  F) = .35

a) P(F  T) = P(F) + P(T) - P(F  T) = .44 + .54 - .35 = .63

b) P(F  T) - P(F  T) = .63 - .35 = .28

c) 1 - P(F  T) = 1 - .63 = .37

d)

Faculty
References

Y
N

Transcript Y .35 .19 .54

N .09 .37 .46

.44 .56 1.00

 P a g e | 164

4.15

C D E F

A 5 11 16 8 40

B 2 3 5 7 17

7 14 21 15 57

a) P(A  E) = 16/57 = .2807

b) P(D  B) = 3/57 = .0526

c) P(D  E) = .0000

d) P(A  B) = .0000

4.16

D E F

A .12 .13 .08 .33

B .18 .09 .04 .31

C .06 .24 .06 .36

.36 .46 .18 1.00

a) P(E  B) = .09

b) P(C  F) = .06

c) P(E  D) = .00

4.17 Let D = Defective part

a) (without replacement)
 P a g e | 165

6 5 30
 
50 49 2450
P(D1  D2) = P(D1)  P(D2  D1) = = .0122

b) (with replacement)

6 6 36
 
50 50 2500
P(D1  D2) = P(D1)  P(D2) = = .0144

4.18 Let U = Urban

I = care for Ill relatives

P(U) = .78 P(I) = .15 P(I U) = .11

a) P(U  I) = P(U)  P(I U)

P(U  I) = (.78)(.11) = .0858

b) P(U  NI) = P(U)  P(NI U) but P(I U) = .11

So, P(NI U) = 1 - .11 = .89 and

P(U  NI) = P(U)  P(NI U) = (.78)(.89) = .6942

c)

Yes No

Yes .15

I No .85
 P a g e | 166

.78 .22

The answer to a) is found in the YES-YES cell. To compute this cell,

take 11%

or .11 of the total (.78) people in urban areas. (.11)(.78) = .0858 which
belongs in

the “YES-YES" cell. The answer to b) is found in the Yes for U and no for
I cell.

It can be determined by taking the marginal, .78, less the answer for
a), .0858.

d. P(NU  I) is found in the no for U column and the yes for I row (1st row and
2nd column). Take the marginal, .15, minus the yes-yes cell, .0858, to get
.0642.
 P a g e | 167

4.19 Let S = stockholder Let C = college

P(S) = .43 P(C) = .37 P(C S) = .75

a) P(NS) = 1 - .43 = .57

b) P(S  C) = P(S) P(C S) = (.43)(.75) = .3225

c) P(S  C) = P(S) + P(C) - P(S  C) = .43 + .37 - .3225 = .4775

d) P(NS  NC) = 1 - P(S  C) = 1 - .4775 = .5225

e) P(NS  NC) = P(NS) + P(NC) - P(NS  NC) = .57 + .63 - .5225 = .

6775

f) P(C  NS) = P(C) - P(C  S) = .37 - .3225 = .0475

The matrix:

Yes No

Yes .3225 .1075 .43

S No .0475 .5225 .57

.37 .63 1.00

 P a g e | 168

4.20 Let F = fax machine Let P = personal computer

Given: P(F) = .10 P(P) = .52 P(P F) = .91

a) P(F  P) = P(F)  P(P F) = (.10)(.91) = .091

b) P(F  P) = P(F) + P(P) - P(F  P) = .10 + .52 - .091 = .529

c) P(F  NP) = P(F)  P(NP F)

Since P(P F) = .91, P(NP F)= 1 - P(P F) = 1 - .91 = .09

P(F  NP) = (.10)(.09) = .009

d) P(NF  NP) = 1 - P(F  P) = 1 - .529 = .471

e) P(NF  P) = P(P) - P(F  P) = .52 - .091 = .429

The matrix:

P NP

F .091 .009 .10

NF .429 .471 .90

.520 .480 1.00

 P a g e | 169

4.21

Let S = safety Let A = age

P(S) = .30 P(A) = .39 P(A S) = .87

a) P(S  NA) = P(S) P(NA S)

but P(NA S) = 1 - P(A S) = 1 - .87 = .13

P(S  NA) = (.30)(.13) = .039

b) P(NS  NA) = 1 - P(S  A) = 1 - [P(S) + P(A) - P(S  A)]

but P(S  A) = P(S)  P(A S) = (.30)(.87) = .261

P(NS  NA) = 1 - (.30 + .39 - .261) = .571

c) P(NS  A) = P(NS) - P(NS  NA)

but P(NS) = 1 - P(S) = 1 - .30 = .70

P(NS  A) = .70 - 571 = .129

The matrix:

Yes No

Yes .261 .039 .30

S No .129 .571 .70

.39 .61 1.00

 P a g e | 170

4.22 Let C = ceiling fans Let O = outdoor grill

P(C) = .60 P(O) = .29 P(C  O) = .13

a) P(C  O)= P(C) + P(O) - P(C  O) = .60 + .29 - .13 = .76

b) P(NC  NO) = 1 - P(C  O)= 1 - .76 = .24

c) P(NC  O) = P(O) - P(C  O) = .29 - .13 = .16

d) P(C  NO) = P(C) - P(C  O) = .60 - .13 = .47

The matrix:

Yes No

Yes .13 .47 .60

No .16 .24 .40

C

.29 .71 1.00

4.23

E F G

A 15 12 8 35

B 11 17 19 47

C 21 32 27 80

D 18 13 12 43

65 74 66 205
 P a g e | 171

a) P(G A) = 8/35 = .2286

b) P(B F) = 17/74 = .2297

c) P(C E) = 21/65 = .3231

d) P(E G) = .0000
 P a g e | 172

4.24

C D

A .36 .44 .80

B .11 .09 .20

.47 .53 1.00

a) P(C A) = .36/.80 = .4500

b) P(B D) = .09/.53 = .1698

c) P(A B) = .0000

4.25

Calculator

Yes No

Yes 46 3 49

No 11 15 26
Computer

57 18 75

Select a category from each variable and test

 P a g e | 173

P(V1 V2) = P(V1).

For example,P(Yes Computer Yes Calculator) = P(Yes Computer)?

46 49

57 75
?

.8070  .6533

Since this is one example that the conditional does not equal
the marginal in

is matrix, the variable, computer, is not independent of the

variable,

calculator.
 P a g e | 174

4.26 Let C = construction Let S = South Atlantic

83,384 total failures

10,867 failures in construction

8,010 failures in South Atlantic

1,258 failures in construction and South Atlantic

a) P(S) = 8,010/83,384 = .09606

b) P(C  S) = P(C) + P(S) - P(C  S) =

10,867/83,384 + 8,010/83,384 - 1,258/83,384 = 17,619/83,384 = .

2113

1258
P (C  S ) 83,384

P(S ) 8010
83,384
c) P(C S) = = .15705

1258
P (C  S ) 83,384

P (C ) 10,867
83,384
d) P(S C) = = .11576

P ( NS  NC ) 1  P (C  S )

P ( NC ) P ( NC )
e) P(NS NC) =

but NC = 83,384 - 10,867 = 72,517

 P a g e | 175

and P(NC) = 72,517/83,384 = .869675

Therefore, P(NS NC) = (1 - .2113)/(.869675) = .9069

P ( NS  C ) P(C )  P (C  S )

P (C ) P (C )
f) P(NS C) =

but P(C) = 10,867/83,384 = .1303

P(C  S) = 1,258/83,384 = .0151

Therefore, P(NS C) = (.1303 - .0151)/.1303 = .8842

 P a g e | 176

4.27 Let E = Economy Let Q = Qualified

P(E) = .46 P(Q) = .37 P(E  Q) = .15

a) P(E Q) = P(E  Q)/P(Q) = .15/.37 = .4054

b) P(Q E) = P(E  Q)/P(E) = .15/.46 = .3261

c) P(Q NE) = P(Q  NE)/P(NE)

but P(Q  NE) = P(Q) - P(Q  E) = .37 - .15 = .22

P(NE) = 1 - P(E) = 1 - .46 = .54

P(Q NE) = .22/.54 = .4074

d) P(NE  NQ) = 1 - P(E  Q) = 1 - [P(E) + P(Q) - P(E  Q)]

= 1 - [.46 + .37 - .15] = 1 - (.68) = .32

The matrix:

Yes No

Yes .15 .31 .46

E No .22 .32 .54

.37 .63 1.00

P a g e | 177
 P a g e | 178

4.28 Let EM = email while on phone Let TD = “to-do” lists during

meetings

P(EM) = .54 P(TD EM) = .20

a) P(EM  TD) = P(EM)  P(TD EM) = (.54)(.20) = .1080

b) P(not TD EM) = 1 - P(TD EM) = 1 - .20 = .80

c) P(not TD  EM) = P(EM) - P(EM  TD) = .54 - .1080 = .4320

Could have been worked as: P(not TD  EM) = P(EM) P(not TD EM)
=

(.54)(.80) = .4320

The matrix:

TD

Yes No

Yes .1080 .4320 .54

No .46
EM

1.00
 P a g e | 179

4.29 Let H = hardware Let S = software

P(H) = .37 P(S) = .54 P(S H) = .97

a) P(NS H) = 1 - P(S H) = 1 - .97 = .03

b) P(S NH) = P(S  NH)/P(NH)

but P(H  S) = P(H)  P(S H) = (.37)(.97) = .3589

so P(NH  S) = P(S) - P(H  S) = .54 - .3589 = .1811

P(NH) = 1 - P(H) = 1 - .37 = .63

P(S NH) = (.1811)/(.63) = .2875

c) P(NH S) = P(NH  S)/P(S) = .1811//54 = .3354

d) P(NH NS) = P(NH  NS)/P(NS)

but P(NH  NS) = P(NH) - P(NH  S) = .63 - .1811 = .4489

and P(NS) = 1 - P(S) = 1 - .54 = .46

P(NH NS) = .4489/.46 = .9759

The matrix:

Yes No

Yes .3589 .0111 .37

 P a g e | 180

No .1811 .4489 .63

H

.54 .46 1.00

 P a g e | 181

4.30 Let R = agreed or strongly agreed that lack of role models was a
barrier

Let S = agreed or strongly agreed that gender-based stereotypes was

a barrier

P(R) = .43 P(S) = .46 P(R S) = .77 P(not S) = .54

a.) P(not R S) = 1 - P(R S) = 1 - .77 = .23

b.) P(not S R) = P(not S  R)/P(R)

but P(S  R) = P(S)  P(R S) = (.46)(.77) = .3542

so P(not S  R) = P(R) - P(S  R) = .43 - .3542 = .0758

Therefore, P(not S R) = (.0758)/(.43) = .1763

c.) P(not R not S) = P(not R  not S)/P(not S)

but P(not R  not S) = P(not S) - P(not S  R) = .54 - .0758 = .

4642

P(not R not S) = .4642/.54 = .8596

The matrix:

Yes No

Yes .3542 .0758 .43

No .1058 .4642 .57

R
 P a g e | 182

.46 .54 1.00

 P a g e | 183

4.31 Let A = product produced on Machine A

B = product produces on Machine B

C = product produced on Machine C

D = defective product

P(A) = .10 P(B) = .40 P(C) = .50

P(D A) = .05 P(D B) = .12 P(D C) = .08

Event Prior Conditional Joint Revised

P(D  Ei)

P(Ei) P(D Ei)

A .10 .05 .005 .005/.093=.0538

B .40 .12 .048 .048/.093=.5161

C .50 .08 .040 .040/.093=.4301

P(D)=.093

Revise: P(A D) = .005/.093 = .0538

P(B D) = .048/.093 = .5161

P(C D) = .040/.093 = .4301

 P a g e | 184

4.32 Let A = Alex fills the order

B = Alicia fills the order

C = Juan fills the order

I = order filled incorrectly

K = order filled correctly

P(A) = .30 P(B) = .45 P(C) = .25

P(I A) = .20 P(I B) = .12 P(I C) = .05

P(K A) = .80 P(K B) = .88 P(K C) = .95

a) P(B) = .45

b) P(K C) = 1 - P(I C) = 1 - .05 = .95

c)

Event Prior Conditional Joint Revised

P(I  Ei)

P(Ei) P(I Ei) P(Ei I)

A .30 .20 .0600 .0600/.1265=.4743

B .45 .12 .0540 .0540/.1265=.4269

C .25 .05 .0125 .0125/.1265=.0988

P(I)=.1265

Revised: P(A I) = .0600/.1265 = .4743

 P a g e | 185

P(B I) = .0540/.1265 = .4269

P(C I) = .0125/.1265 = .0988

d)

Event Prior Conditional Joint Revised

P(K  Ei)

P(Ei) P(K Ei) P(Ei K)

A .30 .80 .2400 .2400/.8735=.2748

B .45 .88 .3960 .3960/.8735=.4533

C .25 .95 .2375 .2375/.8735=.2719

P(K)=.8735
 P a g e | 186

4.33 Let T = lawn treated by Tri-state

G = lawn treated by Green Chem

V = very healthy lawn

N = not very healthy lawn

P(T) = .72 P(G) = .28 P(V T) = .30 P(V G) = .20

Event Prior Conditional Joint Revised

P(V  Ei)

P(Ei) P(V Ei) P(Ei V)

A .72 .30 .216 .216/.272=.7941

B .28 .20 .056 .056/.272=.2059

P(V)=.272

Revised: P(T V) = .216/.272 = .7941

P(G V) = .056/.272 = .2059

4.34 Let S = small Let L = large

The prior probabilities are: P(S) = .70 P(L) = .30

P(T S) = .18 P(T L) = .82

 P a g e | 187

Event Prior Conditional Joint Revised

P(T  Ei)

P(Ei) P(T Ei) P(Ei T)

S .70 .18 .1260 .1260/.3720 = .3387

L .30 .82 .2460 .2460/.3720 = .6613

P(T)=.3720

Revised: P(S T) = .1260/.3720 = .3387

P(L T) = .2460/.3720 = .6613

37.2% offer training since P(T) = .3720,.

4.35

Variable 1

D E

A 10 20 30

Variable 2 B 15 5 20
 P a g e | 188

C 30 15 45

55 40 95

a) P(E) = 40/95 = .42105

b) P(B  D) = P(B) + P(D) - P(B  D)

= 20/95 + 55/95 - 15/95 = 60/95 = .63158

c) P(A  E) = 20/95 = .21053

d) P(B E) = 5/40 = .1250

e) P(A  B) = P(A) + P(B) = 30/95 + 20/95 =

50/95 = .52632

f) P(B  C) = .0000 (mutually exclusive)

g) P(D C) = 30/45 = .66667

P(A  B) .0 0 0 0

P (B ) 20 / 95
h) P(A B) = = .0000 (A and B are mutually
exclusive)

i) P(A) = P(A D)??

Does 30/95 = 10/95 ??

Since, .31579  .18182, Variables 1 and 2 are not independent.

P a g e | 189
 P a g e | 190

D E F G 4.36

A 3 9 7 12 31

B 8 4 6 4 22

C 10 5 3 7 25

21 18 16 23 78

a) P(F  A) = 7/78 = .08974

P (A  B) .0 0 0 0

P (B ) 22 / 78
b) P(A B) = = .0000 (A and B are mutually
exclusive)

c) P(B) = 22/78 = .28205

d) P(E  F) = .0000 Mutually Exclusive

e) P(D B) = 8/22 = .36364

f) P(B D) = 8/21 = .38095

g) P(D  C) = 21/78 + 25/78 – 10/78 = 36/78 = .4615

h) P(F) = 16/78 = .20513

 P a g e | 191

4.37

Age(years)

<35 35-44 45-54 55-64 >65

Male .11 .20 .19 .12 .16 .78

Gender Female .07 .08 .04 .02 .01 .22

.18 .28 .23 .14 .17 1.00

a) P(35-44) = .28

b) P(Woman  45-54) = .04

c) P(Man  35-44) = P(Man) + P(35-44) - P(Man  35-44) = .78 + .28 -

.20 = .86

d) P(<35  55-64) = P(<35) + P(55-64) = .18 + .14 = .32

e) P(Woman 45-54) = P(Woman  45-54)/P(45-54) = .04/.23= .1739

f) P(not W  not 55-64) = .11 + .20 + .19 + .16 = .66

 P a g e | 192

4.38 Let T = thoroughness Let K = knowledge

P(T) = .78 P(K) = .40 P(T  K) = .27

a) P(T  K) = P(T) + P(K) - P(T  K) =

.78 + .40 - .27 = .91

b) P(NT  NK) = 1 - P(T  K) = 1 - .91 = .09

c) P(K T) = P(K  T)/P(T) = .27/.78 = .3462

d) P(NT  K) = P(NT) - P(NT  NK)

but P(NT) = 1 - P(T) = .22

P(NT  K) = .22 - .09 = .13

The matrix:

Yes No

Yes .27 .51 .78

T No .13 .09 .22

.40 .60 1.00

 P a g e | 193

4.39 Let R = retirement Let L = life insurance

P(R) = .42 P(L) = .61 P(R  L) = .33

a) P(R L) = P(R  L)/P(L) = .33/.61 = .5410

b) P(L R) = P(R  L)/P(R) = .33/.42 = .7857

c) P(L  R) = P(L) + P(R) - P(L  R) = .61 + .42 - .33 = .70

d) P(R  NL) = P(R) - P(R  L) = .42 - .33 = .09

e) P(NL R) = P(NL  R)/P(R) = .09/.42 = .2143

The matrix:

Yes No

Yes .33 .09 .42

No .28 .30 .58

R

.61 .39 1.00

 P a g e | 194

4.40 P(T) = .16 P(T W) = .20 P(T NE) = .17

P(W) = .21 P(NE) = .20

a) P(W  T) = P(W)P(T W) = (.21)(.20) = .042

b) P(NE  T) = P(NE)P(T NE) = (.20)(.17) = .034

c) P(W T) = P(W  T)/P(T) = (.042)/(.16) = .2625

d) P(NE NT) = P(NE  NT)/P(NT) = {P(NE)P(NT NE)}/P(NT)

but P(NT NE) = 1 - P(T NE) = 1 - .17 = .83 and

P(NT) = 1 - P(T) = 1 - .16 = .84

Therefore, P(NE NT) = {P(NE)P(NT NE)}/P(NT) =

{(.20)(.83)}/(.84) = .1976

e) P(not W  not NE T) = P(not W  not NE  T)/ P(T)

but P(not W  not NE  T) =

.16 - P(W  T) - P(NE  T) = .16 - .042 - .034 = .084

P(not W  not NE  T)/ P(T) = (.084)/(.16) = .525

 P a g e | 195

4.41 Let M = MasterCard A = American Express V = Visa

P(M) = .30 P(A) = .20 P(V) = .25

P(M  A) = .08 P(V  M) = .12 P(A  V) = .06

a) P(V  A) = P(V) + P(A) - P(V  A) = .25 + .20 - .06 = .39

b) P(V M) = P(V  M)/P(M) = .12/.30 = .40

c) P(M V) = P(V  M)/P(V) = .12/.25 = .48

d) P(V) = P(V M)??

.25  .40

Possession of Visa is not independent of possession of MasterCard

e) American Express is not mutually exclusive of Visa

because P(A  V)  .0000

 P a g e | 196

4.42 Let S = believe SS secure N = don't believe SS will be secure

<45 = under 45 years old >45 = 45 or more years old

P(N) = .51

Therefore, P(S) = 1 - .51 = .49

P(<45) = .57 P(> 45) = .43

P(S >45) = .70

Therefore, P(N >45) = 1 - P(S >45) = 1 - .70 = .30

a) P(>45) = 1 - P(<45) = 1 - .57 = .43

b) P(<45  S) = P(S) - P(>45  S) =

but P(> 45  S) = P(> 45)P(S > 45) = (.43)(.70) = .301

P(<45  S) = P(S) - P(>45  S) = .49 - .301 = .189

c) P(>45 S) = P(>45  S)/P(S) = .189/.49 = .6143

d) (<45  N) = P(<45) + P(N) - P(<45  N) =

but P(<45  N) = P(<45) - P(<45  S) = .57 - .189 = .381

so P(<45  N) = .57 + .51 - .381 = .699

Probability Matrix Solution for Problem 4.42:

 P a g e | 197

S N

<45 .189 .381 .57

>45 .301 .129 .43

.490 .510 1.00

 P a g e | 198

4.43 Let M = expect to save more

R = expect to reduce debt

NM = don't expect to save more

NR = don't expect to reduce debt

P(M) = .43 P(R) = .45 P(R M) = .81

P(NR M) = 1 - P(R M) = 1 - .81 = .19

P(NM) = 1 - P(M) = 1 - .43 = .57

P(NR) = 1 - P(R) = 1 - .45 = .55

a) P(M  R) = P(M)P(R M) = (.43)(.81) = .3483

b) P(M  R) = P(M) + P(R) - P(M  R)

= .43 + .45 - .3483 = .5317

c) P(neither save nor reduce debt) =

1 - P(M  R) = 1 - .5317 = .4683

d) P(M  NR) = P(M)P(NR M) = (.43)(.19) = .0817

Probability matrix for problem 4.43:

Reduce

Yes No

Yes .3483 .0817 .43

Save No .1017 .4683 .57

.45 .55 1.00

P a g e | 199
 P a g e | 200

4.44 Let R = read

Let B = checked in the with boss

P(R) = .40 P(B) = .34 P(B R) = .78

a) P(B  R) = P(R)P(B R) = (.40)(.78) = .312

b) P(NR  NB) = 1 - P(R  B)

but P(R  B) = P(R) + P(B) - P(R  B) =

.40 + .34 - .312 = .428

P(NR  NB) = 1 - .428 = .572

c) P(R B) = P(R  B)/P(B) = (.312)/(.34) = .9176

d) P(NB R) = 1 - P(B R) = 1 - .78 = .22

e) P(NB NR) = P(NB  NR)/P(NR)

but P(NR) = 1 - P(R) = 1 - .40 = .60

P(NB NR) = .572/.60 = .9533

f) Probability matrix for problem 4.44:

B NB

R .312 .088 .40

 P a g e | 201

NR .028 .572 .60

.34 .66 1.00

 P a g e | 202

4.45 Let Q = keep quiet when they see co-worker misconduct

Let C = call in sick when they are well

P(Q) = .35 P(NQ) = 1 - .35 = .65 P(C Q) = .75 P(Q C) = .40

a) P(C  Q) = P(Q)P(C Q) = (.35)(.75) = .2625

b) P(Q  C) = P(Q) + P(C) - P(C  Q)

but P(C) must be solved for:

P(C  Q) = P(C)  P(Q C)

.2625 = P(C) ( .40)

Therefore, P(C) = .2625/.40 = .65625

and P(Q  C) = .35 + .65626 - .2625 = .74375

c) P(NQ C) = P(NQ  C)/P(C)

but P(NQ  C) = P(C) - P(C  Q) = .65625 - .2625 = ..39375

Therefore, P(NQ C) = P(NQ  C)/P(C) = .39375/.65626 = .60

d) P(NQ  NC) = 1 - P(Q  C) = 1 - .74375 = .25625

e) P(Q  NC) = P(Q) - P(Q  C) = .35 - .2625 = .0875

Probability matrix for problem 4.45:

C
 P a g e | 203

Y N

Y .2625 .0875 .35

Q N . . .65
39375 25625

. . 1.00
65625 34375
 P a g e | 204

4.46 Let: D = denial

I = inappropriate

C = customer

P = payment dispute

S = specialty

G = delays getting care

R = prescription drugs

P(D) = .17 P(I) = .14 P(C) = .14 P(P) = .11

P(S) = .10 P(G) = .08 P(R) = .07

a) P(P  S) = P(P) + P(S) = .11 + .10 = .21

b) P(R  C) = .0000 (mutually exclusive)

c) P(I S) = P(I  S)/P(S) = .0000/.10 = .0000

d) P(NG  NP) = 1 - P(G  P) = 1 - [P(G) + P(P)] =

1 – [.08 + .11] = 1 - .19 = .81

 P a g e | 205

4.47 Let R = retention Let P = process improvement

P(R) = .56 P(P  R) = .36 P(R P) = .90

a) P(R  NP) = P(R) - P(P  R) = .56 - .36 = .20

b) P(P R) = P(P  R)/P(R) = .36/.56 = .6429

c) P(P) = ??

Solve P(R P) = P(R  P)/P(P) for P(P):

P(P) = P(R  P)/P(R P) = .36/.90 = .40

d) P(R  P) = P(R) + P(P) - P(R  P) =

.56 + .40 - .36 = .60

e) P(NR  NP) = 1 - P(R  P) = 1 - .60 = .40

f) P(R NP) = P(R  NP)/P(NP)

but P(NP) = 1 - P(P) = 1 - .40 = .60

P(R NP) = .20/.60 = .3333

Y N

Y .36 .20 .56

R N .04 .40 .44

.40 .60 1.00

 P a g e | 206

Note: In constructing the matrix, we are given P(R) = .56, P(P  R)

= .36, and

P(R P) = .90. That is, only one marginal probability is given.

From P(R), we can get P(NR) by taking 1 - .56 = .44.

However, only these two marginal values can be computed

directly.

To solve for P(P), using what is given, since we know that 90%
of P lies

in the intersection and that the intersection is .36, we can set

up an

equation to solve for P:

.90P = .36

Solving for P = .40.

 P a g e | 207

4.48 Let M = mail Let S = sales

P(M) = .38 P(M  S) = .0000 P(NM  NS) = .41

a) P(M  NS) = P(M) - P(M  S) = .38 - .00 = .38

b) Because P(M  S) = .0000, P(M  S) = P(M) + P(S)

Therefore, P(S) = P(M  S) - P(M)

but P(M  S) = 1 - P(NM  NS) = 1 - .41 = .59

Thus, P(S) = P(M  S) - P(M) = .59 - .38 = .21

c) P(S M) = P(S  M)/P(M) = .0000/.38 = .0000

d) P(NM NS) = P(NM  NS)/P(NS) = .41/.79 = .5190

where: P(NS) = 1 - P(S) = 1 - .21 = .79

Probability matrix for problem 4.48:

Y N

Y . .38 .38
0000
M
N .21 .41 .62

.21 .79 1.00

 P a g e | 208

4.49 Let F = Flexible Work Let V = Gives time off for Volunteerism

P(F) = .41 P(V NF) = .10 P(V F) = .60

from this, P(NF) = 1 - .41 = .59

a) P(F  V) = P(F) + P(V) - P(F  V)

P(F) = .41 and P(F  V) = P(F)P(V F) = (.41)(.60) = .246

Find P(V) by using P(V) = P(F  V) + P(NF  V)

but P(NF  V) = P(NF)P(V NF) = (.59)(.10) = .059

so, P(V) = P(F  V) + P(NF  V) = .246 + .059 = .305

and P(F  V) = P(F) + P(V) - P(F  V) = .41 + .305 - .246 = .469

b) P(F  NV) = P(F) - P(F  V) = .41 - .246 = .164

c) P(F NV) = P(F  NV)/P(NV)

P(F  NV) = .164

P(NV) = 1 – P(V) = 1 - .305 = .695.

P(F NV) = P(F  NV)/P(NV) = .164/.695 = .2360

d) P(NF V) = P(NF  V)/P(V) = .059/.305 = .1934

e) P(NF  NV) = P(NF) + P(NV) - P(NF  NV)

P(NF) = .59 P(NV) = .695

Solve for P(NF  NV) = P(NV) – P(F  NV) = .695 - .164 = .531

P(NF  NV) = P(NF) + P(NV) - P(NF  NV) = .59 + .695 - .531 = .

754

Probability matrix for problem 4.49:

V
 P a g e | 209

Y N

Y .246 .164 .41

F N .059 .531 .59

.305 .695 1.00

 P a g e | 210

4.50 Let S = Sarabia Let T = Tran Let J = Jackson Let B =

blood test

P(S) = .41 P(T) = .32 P(J) = .27

P(B S) = .05 P(B T) = .08 P(B J) = .06

Event Prior Conditional Joint Revised

P(B  Ei)

P(Ei) P(B Ei) P(Bi NS)

S .41 .05 .0205 .3291

T .32 .08 .0256 .4109

J .27 .06 .0162 .2600

P(B) = .0623

4.51 Let R = regulations T = tax burden

P(R) = .30 P(T) = .35 P(T R) = .71

a) P(R  T) = P(R)P(T R) = (.30)(.71) = .2130

b) P(R  T) = P(R) + P(T) - P(R  T) =

.30 + .35 - .2130 = .4370

c) P(R  T) - P(R  T) = .4370 - .2130 = .2240

d) P(R T) = P(R  T)/P(T) = .2130/.35 = .6086

 P a g e | 211

e) P(NR T) = 1 - P(R T) = 1 - .6086 = .3914

f) P(NR NT) = P(NR  NT)/P(NT) = [1 - P(R  T)]/P(NT) =

(1 - .4370)/.65 = .8662

Probability matrix for problem 4.51:

Y N

Y .213 .087 .30

R N .137 .563 .70

.35 .65 1.00

 P a g e | 212

4.52

Event Prior Conditional Joint Revised

P(RB  Ei)

P(Ei) P(RB Ei) P(Ei RB)

0-24 .353 .11 .03883 .03883/.25066 = .

15491

25-34 .142 .24 .03408 .03408/.25066 = .

13596

35-44 .160 .27 .04320 .04320/.25066 = .

17235

> 45 .345 .39 .13455 .13455/.25066 = .

53678

P(RB) = .
25066

4.53 Let GH = Good health Let HM = Happy marriage Let FG =

Faith in God

P(GH) = .29 P(HM) = .21 P(FG) = .40

a) P(HM  FG) = P(HM) + P(FG) - P(HM  FG)

but P(HM  FG) = .0000

P(HM  FG) = P(HM) + P(FG) = .21 + .40 = .61

b) P(HM  FG  GH) = P(HM) + P(FG) + P(GH) =

 P a g e | 213

.29 + .21 + .40 = .9000

c) P(FG  GH) = .0000

The categories are mutually exclusive.

The respondent could not select more than one answer.

d) P(neither FG nor GH nor HM) = 1 - P(HM  FG  GH) = 1 - .9000 =

.1000
 P a g e | 214

Chapter 5
Discrete Distributions

LEARNING OBJECTIVES

The overall learning objective of Chapter 5 is to help you understand a

category of probability distributions that produces only discrete outcomes,
thereby enabling you to:

1. Distinguish between discrete random variables and continuous random

variables.

2. Know how to determine the mean and variance of a discrete

distribution.

3. Identify the type of statistical experiments that can be described by

the binomial distribution and know how to work such problems.

4. Decide when to use the Poisson distribution in analyzing statistical

experiments and know how to work such problems.

5. Decide when binomial distribution problems can be approximated by

the Poisson distribution and know how to work such problems.

6. Decide when to use the hypergeometric distribution and know how to

work such problems

CHAPTER TEACHING STRATEGY

 P a g e | 215

Chapters 5 and 6 introduce the student to several statistical

distributions. It is important to differentiate between the discrete
distributions of chapter 5 and the continuous distributions of chapter 6.

The approach taken in presenting the binomial distribution is to build

on techniques presented in chapter 4. It can be helpful to take the time to
apply the law of multiplication for independent events to a problem and
demonstrate to students that sequence is important. From there, the student
will more easily understand that by using combinations, one can more quickly
determine the number of sequences and weigh the probability of obtaining a
single sequence by that number. In a sense, we are developing the binomial
formula through an inductive process. Thus, the binomial formula becomes
more of a summary device than a statistical "trick". The binomial tables
presented in this text are non cumulative. This makes it easier for the
student to recognize that the table is but a listing of a series of binomial
formula computations. In addition, it lends itself more readily to the graphing
of a binomial distribution.

It is important to differentiate applications of the Poisson distribution

from binomial distribution problems. It is often difficult for students to
determine which type of distribution to apply to a problem. The Poisson
distribution applies to rare occurrences over some interval. The parameters
involved in the binomial distribution (n and p) are different from the
parameter (Lambda) of a Poisson distribution.

It is sometimes difficult for students to know how to handle Poisson

problems in which the interval for the problem is different than the stated
interval for Lambda. Note that in such problems, it is always the value of
Lambda that is adjusted not the value of x. Lambda is a

long-run average that can be appropriately adjusted for various intervals. For
example, if a store is averaging  customers in 5 minutes, then it will also be
averaging 2 customers in 10 minutes. On the other hand, x is a one-time
observation and just because x customers arrive in 5 minutes does not mean
that 2x customers will arrive in 10 minutes.

Solving for the mean and standard deviation of binomial distributions

prepares the students for chapter 6 where the normal distribution is
sometimes used to approximate binomial distribution problems. Graphing
binomial and Poisson distributions affords the student the opportunity to
visualize the meaning and impact of a particular set of parameters for a
distribution. In addition, it is possible to visualize how the binomial
 P a g e | 216

distribution approaches the normal curve as p gets nearer to .50 and as n

gets larger for other values of p. It can be useful to demonstrate this in class
along with showing how the graphs of Poisson distributions also approach the
normal curve as  gets larger.

In this text (as in most) because of the number of variables used in its
computation, only exact probabilities are determined for hypergeometric
distribution. This, combined with the fact that there are no hypergeometric
tables given in the text, makes it cumbersome to determine cumulative
probabilities for the hypergeometric distribution. Thus, the hypergeometric
distribution can be presented as a fall-back position to be used only when the
binomial distribution should not be applied because of the non independence
of trials and size of sample.
 P a g e | 217

CHAPTER OUTLINE

5.1 Discrete Versus Continuous Distributions

5.2 Describing a Discrete Distribution

Mean, Variance, and Standard Deviation of Discrete

Distributions

Mean or Expected Value

Variance and Standard Deviation of a Discrete Distribution

5.3 Binomial Distribution

Solving a Binomial Problem

Using the Binomial Table

Using the Computer to Produce a Binomial Distribution

Mean and Standard Deviation of the Binomial Distribution

Graphing Binomial Distributions

5.4 Poisson Distribution

Working Poisson Problems by Formula

Using the Poisson Tables

Mean and Standard Deviation of a Poisson Distribution

Graphing Poisson Distributions

Using the Computer to Generate Poisson Distributions

Approximating Binomial Problems by the Poisson Distribution

 P a g e | 218

5.5 Hypergeometric Distribution

Using the Computer to Solve for Hypergeometric Distribution

Probabilities

KEY TERMS

Binomial Distribution Hypergeometric

Distribution

Continuous Distributions Lambda ()

Continuous Random Variables Mean, or Expected Value

Discrete Distributions Poisson

Distribution

Discrete Random Variables Random Variable

 P a g e | 219

SOLUTIONS TO PROBLEMS IN CHAPTER 5

5.1 x P(x) x·P(x) (x-µ)2

(x-µ)2·P(x)

1 .238 .238 2.775556

0.6605823

2 .290 .580 0.443556

0.1286312

3 .177 .531 0.111556

0.0197454

4 .158 .632 1.779556

0.2811700

5 .137 .685 5.447556

0.7463152

µ = [x·P(x)] = 2.666 2 = [(x-

µ)2·P(x)] = 1.836444

1.836444
 = = 1.355155

5.2 x P(x) x·P(x) (x-µ)2

2
(x-µ) ·P(x)

0 .103
.000 7.573504
0.780071
1 .118 .118 3.069504
0.362201

2 .246 .492 0.565504

0.139114
 P a g e | 220

3 .229 .687 0.061504

0.014084

4 .138 .552 1.557504

0.214936

5 .094 .470 5.053504

0.475029

6 .071 .426 10.549500

0.749015

7 .001 .007 18.045500

0.018046

µ =  [x·P(x)] = 2.752 2 = [(x-

µ)2·P(x)] = 2.752496

2.752496
 = = 1.6591

5.3 x P(x) x·P(x) (x-µ)2

2
(x-µ) ·P(x)

0 .461 .000 0.913936

0.421324

1 .285 .285 0.001936

0.000552

2 .129 .258 1.089936

0.140602

3 .087 .261 4.177936

0.363480

4 .038 .152 9.265936

0.352106

E(x) = µ = [x·P(x)]= 0.956 2 = [(x-

2
µ) ·P(x)] = 1.278064
 P a g e | 221

1.278064
 = = 1.1305
 P a g e | 222

5.4 x P(x) x·P(x) (x-µ)2 (x-

2
µ) ·P(x)

0 .262 .000 1.4424

0.37791

1 .393 .393 0.0404

0.01588

2 .246 .492 0.6384

0.15705

3 .082 .246 3.2364

0.26538

4 .015 .060 7.8344

0.11752

5 .002 .010 14.4324

0.02886

6 .000 .000 23.0304

0.00000

µ =  [x·P(x)] = 1.201 2 =  [(x-µ)2·P(x)]

= 0.96260

.96260
 = = .98112

5.5 a) n=4 p = .10 q = .90

P(x=3) = 4 C3(.10)3(.90)1 = 4(.001)(.90) = .0036

b) n=7 p = .80 q = .20

P(x=4) = 7 C4(.80)4(.20)3 = 35(.4096)(.008) = .1147

 P a g e | 223

c) n = 10 p = .60 q = .40

P(x > 7) = P(x=7) + P(x=8) + P(x=9) + P(x=10) =

10C7(.60)7(.40)3 + 10 C8(.60)8(.40)2 + C9(.60)9(.40)1

10

+10C10(.60)10(.40)0 =

120(.0280)(.064) + 45(.0168)(.16) + 10(.0101)(.40) + 1(.0060)

(1) =

.2150 + .1209 + .0403 + .0060 = .3822

d) n = 12 p = .45 q = .55

P(5 < x < 7) = P(x=5) + P(x=6) + P(x=7) =

12C5(.45)5(.55)7 + 12 C6(.45)6(.55)6 + C7(.45)7(.55)5 =

12

792(.0185)(.0152) + 924(.0083)(.0277) + 792(.0037)(.0503) =

.2225 + .2124 + .1489 = .5838

5.6 By Table A.2:

 P a g e | 224

a) n = 20 p = .50

P(x=12) = .120

b) n = 20 p = .30

P(x > 8) = P(x=9) + P(x=10) + P(x=11) + ...+ P(x=20) =

.065 + .031 + .012 + .004 + .001 + .000 = .113

c) n = 20 p = .70

P(x < 12) = P(x=11) + P(x=10) + P(x=9) + ... + P(x=0)

=

.065 + .031 + .012 + .004 + .001 + .000 = .113

d) n = 20 p = .90

P(x < 16) = P(x=16) + P(x=15) + P(x=14) + ...+ P(x=0) =

.090 + .032 + .009 + .002 + .000 = .133

 P a g e | 225

e) n = 15 p = .40

P(4 < x < 9) =

P(x=4) + P(x=5) + P(x=6) + P(x=7) + P(x=8) + P(x=9) =

.127 + .186 + .207 + .177 + .118 + .061 = .876

f) n = 10 p = .60

P(x > 7) = P(x=7) + P(x=8) + P(x=9) + P(x=10) =

.215 + .121 + .040 + .006 = .382

 P a g e | 226

5.7 a) n = 20
p = .70
q = .30

µ = np = 20(.70) = 14

n  p  q  20(.70)(.30)  4.2
 = = 2.05

b) n = 70 p = .35 q = .65

µ = np = 70(.35) = 24.5

n  p  q  70(.35)(.65)  15.925
 = = 3.99

c) n = 100 p = .50 q = .50

µ = np = 100(.50) = 50

n  p  q  100(.50)(. 50)  25
 = = 5

5.8 a) n=6 p = .70 x Prob

0 .001
 P a g e | 227

1 .010

2 .060
3 .185

4 .324
5 .303

6 .118

b) n = 20 p = .50 x Prob

0 .000

1 .000

2 .000
3 .001

4 .005

5 .015

6 .037

7 .074

8 .120

9 .160
 P a g e | 228

10 .176

11 .160

12 .120

13 .074

14 .037

15 .015

16 .005

17 .001

18 .000

19 .000

20 .000
 P a g e | 229

c) n=8 p = .80 x Prob

0 .000

1 .000

2 .001

3 .009

4 .046

5 .147

6 .294

7 .336

8 .168

5.9 a) n = 20 p = .78 x = 14

C14 (.78)14(.22)6 = 38,760(.030855)(.00011338) = .1356

20
 P a g e | 230

b) n = 20 p = .75 x = 20

C20 (.75)20(.25)0 = (1)(.0031712)(1) = .0032

20

c) n = 20 p = .70 x < 12

Use table A.2:

P(x=0) + P(x=1) + . . . + P(x=11)=

.000 + .000 + .000 + .000 + .000 + .000 + .000 +

.001 + .004 + .012 + .031 + .065 = .113

 P a g e | 231

5.10 n = 16 p = .40

P(x > 9): from Table A.2:

x Prob

9 .084

10 .039

11 .014

12 .004

13 .001

.142

P(3 < x < 6):

x Prob

3 .047

4 .101

5 .162

6 .198

.508

n = 13 p = .88

P(x = 10) = 13C10(.88)10(.12)3 = 286(.278500976)(.001728) = .

1376
 P a g e | 232

P(x = 13) = C13(.88)13(.12)0 = (1)(.1897906171)(1) = .1898

13

Expected Value = µ = np = 13(.88) = 11.44

 P a g e | 233

5.11 n = 25 p = .60

a) x > 15

P(x > 15) = P(x = 15) + P(x = 16) + · · · + P(x = 25)

Using Table A.2 n = 25, p = .60

x Prob

15 .161
16 .151
17 .120

18 .080

19 .044

20 .020

21 .007

22 .002

.585

b) x > 20

P(x > 20) = P(x = 21) + P(x = 22) + P(x = 23) + P(x = 24) +
P(x = 25) =

Using Table A.2 n = 25, p = .60

.007 + .002 + .000 + .000 + .000 = .009

 P a g e | 234

c) P(x < 10)

Using Table A.2 n = 25, p = .60 and x = 0, 1, 2, 3, 4, 5, 6, 7,

8, 9

x Prob.

9 .009
8 .003

7 .001

<6 .000

.013
 P a g e | 235

5.12 n = 16 p = .50 x > 10

Using Table A.2, n = 16 and p = .50, P(x=11) + P(x=12) + . . . +

P(x=16) =

x Prob.

11 .067
12 .028

13 .009

14 .002

15 .000
16 .000
.106

For n = 10 p = .87 x=6

C6 (.87)6(.13)4 = 210(.433626)(.00028561) = .0260

10

5.13 n = 15 p = .20

a) P(x = 5) = 15 C5(.20)5(.80)10 = 3003(.00032)(.1073742) = .1032

b) P(x > 9): Using Table A.2

P(x = 10) + P(x = 11) + . . . + P(x = 15) = .000 + .000 + . . . + .

000 = .000
 P a g e | 236

c) P(x = 0) = 15 C0(.20)0(.80)15 = (1)(1)(.035184) = .0352

d) P(4 < x < 7): Using Table A.2

P(x = 4) + P(x = 5) + P(x = 6) + P(x = 7) = .188 + .103 + .043

+ .014 = .348

e)

5.14 n = 18

a) p =.30 µ = 18(.30) = 5.4

p = .34 µ = 18(.34) = 6.12

b) P(x > 8) n = 18 p = .30

from Table A.2

x Prob
 P a g e | 237

8 .081

9 .039

10 .015

11 .005

12 .001

.141

c) n = 18 p = .34

P(2 < x < 4) = P(x = 2) + P(x = 3) + P(x = 4) =

C2(.34)2(.66)16 +
18 C3(.34)3(.66)15 +
18 C4(.34)4(.66)14 =
18

.0229 + .0630 + .1217 = .2076

d) n = 18 p = .30 x=0

C0(.30)0(.70)18 = .00163
18

n = 18 p = .34 x=0

C0(.34)0(.66)18 = .00056
18

Since only 30% (compared to 34%) fall in the $500,000 to

$1,000,000 category, it is

more likely that none of the CPA financial advisors would fall in this
category.
 P a g e | 238

 2 .3
2 .3 5  e ( 6 4 . 3 6 3 4 3 ) ( .1 0 0 2 5 9 )

5! 120
5.15 a) P(x=5 = 2.3) = = .0538

 3 .9
3 .9 2  e ( 1 5 .2 1 ) ( .0 2 0 2 4 2 )

2! 2
b) P(x=2 = 3.9) = = .1539

c) P(x < 3 = 4.1) = P(x=3) + P(x=2) + P(x=1) + P(x=0) =

 4 .1
4 .1 3  e ( 6 8 .9 2 1 ) ( .0 1 6 5 7 3 )

3! 6
= .1904

 4 .1
4 .1 2  e ( 1 6 .8 1 ) ( .0 1 6 5 7 3 )

2! 2
= .1393

 4 .1
4 .1 1  e ( 4 .1 ) ( . 0 1 6 5 7 3 )

1! 1
= .0679

 4 .1
4 .1 0  e ( 1 ) ( .0 1 6 5 7 3 )

0! 1
= .0166

.1904 + .1393 + .0679 + .0166 = .4142

d) P(x=0 = 2.7) =
 P a g e | 239

 2 .7
2 .7 0  e ( 1 ) ( .0 6 7 2 1 )

0! 1
= .0672

e) P(x=1  = 5.4)=

 5 .4
5 .4 1  e ( 5 .4 ) ( .0 0 4 5 1 6 6 )

1! 1
= .0244

f) P(4 < x < 8  = 4.4): P(x=5 = 4.4) + P(x=6 = 4.4) +

P(x=7 = 4.4)=

 4 .4  4 .4  4 .4
4 .4 5  e 4 .4 6  e 4 .4 7  e
5! 6! 7!
+ + =

( 1 6 4 9 .1 6 2 2 ) ( . 0 1 2 2 7 7 3 4 ) ( 7 2 5 6 .3 1 3 9 ) ( .0 1 2 2 7 7 3 4 )
120 720
+ +
( 3 1 , 9 2 7 .7 8 1 ) ( .0 1 2 2 7 7 3 4 )
5040

= .1687 + .1237 + .0778 = .3702

5.16 a) P(x=6 = 3.8) = .0936

b) P(x>7 = 2.9):

x Prob

8 .0068

9 .0022

10 .0006
 P a g e | 240

11 .0002

12 .0000

.0098

c) P(3 < x < 9 = 4.2)=

x Prob

3 .1852

4 .1944

5 .1633
6 .1143
7 .0686

8 .0360

9 .0168

.7786

d) P(x=0 = 1.9) = .1496

e) P(x < 6 = 2.9)=

x Prob

0 .0550

1 .1596

2 .2314

3 .2237

4 .1622

5 .0940
 P a g e | 241

6 .0455

.9714

f) P(5 < x < 8  = 5.7) =

x Prob

6 .1594
7 .1298

8 .0925

.3817

6.3
5.17 a)  = 6.3 mean = 6.3 Standard deviation = = 2.51

x Prob

0 .0018
1 .0116

2 .0364

3 .0765

4 .1205

5 .1519

6 .1595

7 .1435

8 .1130

9 .0791

10 .0498
 P a g e | 242

11 .0285

12 .0150

13 .0073

14 .0033

15 .0014

16 .0005

17 .0002

18 .0001

19 .0000
 P a g e | 243

1.3
b)  = 1.3 mean = 1.3 standard deviation = = 1.14

x Prob

0 .2725
1 .3542

2 .2303

3 .0998

4 .0324

5 .0084

6 .0018

7 .0003

8 .0001

9 .0000
 P a g e | 244

c)  = 8.9 mean = 8.9 standard deviation

8.9
= = 2.98

x Prob

0 .0001
1 .0012

2 .0054

3 .0160

4 .0357

5 .0635

6 .0941

7 .1197

8 .1332

9 .1317

10 .1172

11 .0948

12 .0703

13 .0481

14 .0306

15 .0182

16 .0101

17 .0053

18 .0026

19 .0012

20 .0005

21 .0002
 P a g e | 245

22 .0001
 P a g e | 246

0.6
d)  = 0.6 mean = 0.6 standard deviation = = .775

x Prob

0 .5488

1 .3293
2 .0988

3 .0198

4 .0030

5 .0004

6 .0000
 P a g e | 247

5.18  = 2.8 4 minutes

a) P(x=6  = 2.8)

from Table A.3 .0407

b) P(x=0  = 2.8) =

from Table A.3 .0608

c) Unable to meet demand if x > 4 4 minutes:

x Prob.

5 .0872

6 .0407

7 .0163

8 .0057

9 .0018

10 .0005

11 .0001

.1523

There is a .1523 probability of being unable to meet the

demand.

Probability of meeting the demand = 1 - (.1523) = .8477

 P a g e | 248

15.23% of the time a second window will need to be opened.

d)  = 2.8 arrivals 4 minutes

P(x=3) arrivals 2 minutes = ??

Lambda must be changed to the same interval (½ the size)

New lambda=1.4 arrivals 2 minutes

P(x=3) =1.4) = from Table A.3 = .1128

P(x > 5 8 minutes) = ??

Lambda must be changed to the same interval(twice the size):

New lambda = 5.6 arrivals 8 minutes

 P a g e | 249

P(x > 5   = 5.6):

From Table A.3: x Prob.

5 .1697

6 .1584
7 .1267

8 .0887

9 .0552

10 .0309

11 .0157

12 .0073

13 .0032

14 .0013

15 .0005

16 .0002

17 .0001

.6579

5.19  = x/n = 126/36 = 3.5

Using Table A.3

a) P(x = 0) = .0302

b) P(x > 6) = P(x = 6) + P(x = 7) + . . . =

.0771 + .0385 + .0169 + .0066 + .0023 +

 P a g e | 250

.0007 + .0002 + .0001 = .1424

c) P(x < 4  10 minutes)

Double Lambda to  = 7.0 10 minutes

P(x < 4) = P(x = 0) + P(x = 1) + P(x = 2) + P(x = 3) =

.0009 + .0064 + .0223 + .0521 = .0817

d) P(3 < x < 6 10 minutes)

 = 7.0  10 minutes

P(3 < x < 6) = P(x = 3) + P(x = 4) + P(x = 5) + P(x = 6)

= .0521 + .0912 + .1277 + .1490 = .42

e) P(x = 8  15 minutes)

Change Lambda for a 15 minute interval by multiplying the original

Lambda by 3.

 = 10.5  15 minutes
 P a g e | 251

 x  e   (10.58 )(e 10.5 )


x! 8!
P(x = 8 15 minutes) = = .1009

5.20  = 5.6 days 3 weeks

a) P(x=0  = 5.6):

from Table A.3 = .0037

b) P(x=6 
 = 5.6):

from Table A.3 = .1584

c) P(x > 15 = 5.6):

x Prob.

15 .0005

16 .0002

17 .0001

.0008

Because this probability is so low, if it actually occurred, the

researcher would

question the Lambda value as too low for this period. Perhaps the
value of
 P a g e | 252

Lambda has changed because of an overall increase in pollution.

 P a g e | 253

5.21  = 0.6 trips 1 year

a) P(x=0  = 0.6):

from Table A.3 = .5488

b) P(x=1 
 = 0.6):

from Table A.3 = .3293

c) P(x > 2  = 0.6):

from Table A.3 x Prob.

2 .0988

3 .0198

4 .0030

5 .0004
6 .0000

.1220

d) P(x < 3  3 year period):

The interval length has been increased (3 times)

New Lambda =  = 1.8 trips 3 years

 P a g e | 254

P(x < 3 
 = 1.8):

from Table A.3 x Prob.

0 .1653

1 .2975

2 .2678

3 .1607

.8913

e) P(x=4 6 years):

The interval has been increased (6 times)

New Lambda =  = 3.6 trips 6 years

P(x=4 = 3.6):

from Table A.3 = .1912

5.22  = 1.2 collisions 4 months

a) P(x=0  = 1.2):

from Table A.3 = .3012

 P a g e | 255

b) P(x=2 2 months):

The interval has been decreased (by ½)

New Lambda =  = 0.6 collisions 2 months

P(x=2 
 = 0.6):

from Table A.3 = .0988

c) P(x < 1 collision 6 months):

The interval length has been increased (by 1.5)

New Lambda =  = 1.8 collisions 6 months

P(x < 1  = 1.8):

from Table A.3 x Prob.

0 .1653

1 .2975

.4628
 P a g e | 256

The result is likely to happen almost half the time (46.26%). Ship
channel and

weather conditions are about normal for this period. Safety

awareness is

about normal for this period. There is no compelling reason to

reject the

lambda value of 0.6 collisions per 4 months based on an outcome of

0 or 1

collisions per 6 months.

 P a g e | 257

5.23  = 1.2 pens carton

a) P(x=0   = 1.2):

from Table A.3 = .3012

b) P(x > 8   = 1.2):

from Table A.3 = .0000

c) P(x > 3   = 1.2):

from Table A.3 x Prob.

4 .0260

5 .0062

6 .0012

7 .0002
8 .0000

.0336
 P a g e | 258

5.24 n = 100,000 p = .00004

P(x > 7 n = 100,000 p = .00004):

 = µ = np = 100,000(.00004) = 4.0

Since n > 20 and np < 7, the Poisson approximation to this binomial
problem is

close enough.

P(x > 7   = 4):

Using Table A.3 x Prob.

7 .0595

8 .0298

9 .0132

10 .0053

11 .0019

12 .0006

13 .0002

14 .0001

.1106

P(x >10   = 4):

Using Table A.3 x Prob.

11 .0019
 P a g e | 259

12 .0006

13 .0002

14 .0001

.0028

Since getting more than 10 is a rare occurrence, this particular

geographic region

appears to have a higher average rate than other regions. An

investigation of

particular characteristics of this region might be warranted.

 P a g e | 260

5.25 p = .009 n = 200

Use the Poisson Distribution:

 = np = 200(.009) = 1.8

a) P(x > 6) from Table A.3 =

P(x = 6) + P(x = 7) + P(x = 8) + P(x = 9) + . . . =

.0078 + .0020 + .0005 + .0001 = .0104

b) P(x > 10) = .0000

c) P(x = 0) = .1653

d) P(x < 5) = P(x = 0) + P(x = 1) + P(x = 2) + P( x = 3) + P(x = 4) =

.1653 + .2975 + .2678 + .1607 + .0723 = .9636

5.26 If 99% see a doctor, then 1% do not see a doctor. Thus, p = .01 for
this problem.
 P a g e | 261

n = 300, p = .01,  = n(p) = 300(.01) = 3

a) P(x = 5):

Using  = 3 and Table A.3 = .1008

b) P(x < 4) = P(x = 0) + P(x = 1) + P(x = 2) + P(x = 3) =

.0498 + .1494 + .2240 + .2240 = .6472

c) The expected number = µ =  = 3

 P a g e | 262

5.27 a) P(x = 3  N = 11, A = 8, n = 4)

C 3 3 C1 (56)(3)
8

11 C 4 330
= .5091

b) P(x < 2) N = 15, A = 5, n = 6)

P(x = 1) + P (x = 0) =

5 C1 10 C 5 5 C 0 10 C 6 (5)(252) (1)( 210)


15 C 6 15 C 6 5005 5005
+ =

.2517 + .0420 = .2937

c) P(x=0  N = 9, A = 2, n = 3)

C 0  7 C 3 (1)(35)
2

9 C3 84
= .4167

d) P(x > 4  N = 20, A = 5, n = 7) =

P(x = 5) + P(x = 6) + P(x = 7) =

 P a g e | 263

5 C 5 15 C 2 5 C 6 15 C1 5 C 7 15 C 0

20 C 7 20 C 7 20 C 7
+ + =

(1)(105)
77520
+ 5C6 (impossible) + 5C7(impossible) = .0014
 P a g e | 264

5.28 N = 19 n = 6

a) P(x = 1 private) A = 11

C1 8 C 5 (11)(56)
11

19 C 6 27,132
= .0227

b) P(x = 4 private)

C 4 8 C 2 (330)( 28)
11

19 C 6 27,132
= .3406

c) P(x = 6 private)

C 6 8 C 0 (462)(1)
11

19 C 6 27,132
= .0170

d) P(x = 0 private)

C 0 8 C 6 (1)( 28)
11

19 C 6 27,132
= .0010
 P a g e | 265

5.29 N = 17 A=8 n=4

8 C 0 9 C 4 (1)(126)
17 C 4 2380
a) P(x = 0) = = = .0529

8 C 4 9 C 0 (70)(1)
17 C 4 2380
b) P(x = 4) = = = .0294

9 C 2 8 C 2 (36)( 28)
17 C 4 2380
c) P(x = 2 non computer) = = = .4235

5.30 N = 20 A = 16 white N - A = 4 red n=5

16 C 4  4 C1 (1820)( 4)
20 C 5 15504
a) P(x = 4 white) = = = .4696

4 C 4 16 C1 (1)(16)
20 C 5 15504
b) P(x = 4 red) = = = .0010
 P a g e | 266

4 C 5 16 C 0
20 C 5
c) P(x = 5 red) = = .0000 because 4C5 is impossible to
determine

The participant cannot draw 5 red beads if there are only 4 to draw
from.
 P a g e | 267

5.31 N = 10 n=4

a) A = 3 x = 2

C 2  7 C 2 (3)( 21)
3

C
10 4 210
P(x = 2) = = .30

b) A = 5 x = 0

C 0 5 C 4 (1)(5)
5

10 C 4 210
P(x = 0) = = .0238

c) A = 5 x = 3

C 3 5 C1 (10)(5)
5

C
10 4 210
P(x = 3) = = .2381

5.32 N = 16 A = 4 defective n=3

C 0 12 C 3 (1)( 220)

4

C
16 3 560
a) P(x = 0) = = .3929
 P a g e | 268

C 3 12 C 0 (4)(1)
4

16 C 3 560
b) P(x = 3) = = .0071

4 C 2 12 C1
16 C 3
c) P(x > 2) = P(x=2) + P(x=3) = + .0071 (from part b.)
=

(6)(12)
560
+ .0071 = .1286 + .0071 = .1357

d) P(x < 1) = P(x=1) + P(x=0) =

4 C1 12 C 2 (4)( 66)

16 C 3 560
+ .3929 (from part a.) = + .3929 = .4714 + .
3929 = .8643
 P a g e | 269

5.33 N = 18 A = 11 Hispanic n=5

P(x < 1) = P(1) + P(0) =

11 C1  7 C 4 11 C 0 7 C 5 (11)(35) (1)( 21)


18 C 5 18 C 5 8568 8568
+ = = .0449 + .0025 = .
0474

It is fairly unlikely that these results occur by chance. A researcher

might want to

further investigate this result to determine causes. Were officers

selected based on

leadership, years of service, dedication, prejudice, or some other

reason?

5.34 a) P(x=4  n = 11 and p = .23)

11 C4(.23)4(.77)7 = 330(.0028)(.1605) = .1482

b) P(x > 1 n = 6 and p = .50) =

1 - P(x < 1) = 1 - P(x = 0) =

1 – [6C0(.50)0(.50)6] = 1 – [(1)(1)(.0156)] = .9844

 P a g e | 270

c) P(x > 7  n = 9 and p = .85) = P(x = 8) + P(x = 9) =

9 C8(.85)8(.15)1 + 9C9(.85)9(.15)0 =

(9)(.2725)(.15) + (1)(.2316)(1) = .3679 + .2316 = .5995

d) P(x < 3  n = 14 and p = .70) =

P(x = 3) + P(x = 2) + P(x = 1) + P(x = 0) =

14 C3(.70)3(.30)11 + C2(.70)2(.30)12 +
14

14 C1(.70)1(.30)13 + C0(.70)0(.30)14 =
14

(364)(.3430)(.00000177) + (91)(.49)(.000000531)=

(14)(.70)(.00000016) + (1)(1)(.000000048) =

.0002 + .0000 + .0000 + .0000 = .0002

 P a g e | 271

5.35 a) P(x = 14  n = 20 and p = .60) = .124

b) P(x < 5  n = 10 and p =.30) =

P(x = 4) + P(x = 3) + P(x = 2) + P(x = 1) + P(x=0) =

x Prob.

0 .028

1 .121

2 .233

3 .267

4 .200

.849

c) P(x > 12  n = 15 and p = .60) =

P(x = 12) + P(x = 13) + P(x = 14) + P(x = 15)

x Prob.

12 .063
13 .022

14 .005

15 .000

.090
 P a g e | 272

d) P(x > 20  n = 25 and p = .40) = P(x = 21) + P(x = 22) +

P(x = 23) + P(x = 24) + P(x=25) =

x Prob.

21 .000

22 .000

23 .000
24 .000

25 .000

.000
 P a g e | 273

5.36

a) P(x = 4 = 1.25)

 1 .2 5
( 1 .2 5 4 ) ( e ) ( 2 .4 4 1 4 ) ( .2 8 6 5 )

4! 24
= .0291

b) P(x < 1 = 6.37) = P(x = 1) + P(x = 0) =

 6 .3 7  6 .3 7
( 6 .3 7 ) 1 ( e ) ( 6 .3 7 0 ) ( e ) ( 6 .3 7 ) ( .0 0 1 7 ) ( 1 ) ( .0 0 1 7 )
  
1! 0! 1 1

.0109 + .0017 = .0126

c) P(x > 5 = 2.4) = P(x = 6) + P(x = 7) + ... =

 2 .4  2 .4  2 .4  2 .4  2 .4
( 2 .4 6 ) ( e ) ( 2 .4 7 ) ( e ) ( 2 .4 8 ) ( e ) ( 2 .4 9 ) ( e ) ( 2 .4 10
)(e )
     
6! 7! 8! 9! 10!

.0241 + .0083 + .0025 + .0007 + .0002 = .0358

for values x > 11 the probabilities are each .0000 when rounded off
to 4

decimal places.
 P a g e | 274

5.37 a) P(x = 3 = 1.8) = .1607

b) P(x < 5 = 3.3) =

P(x = 4) + P(x = 3) + P(x = 2) + P(x = 1) + P(x = 0) =

x Prob.

0 .0369

1 .1217

2 .2008

3 .2209

4 .1823

.7626

c) P(x > 3  = 2.1) =

x Prob.

3 .1890

4 .0992

5 .0417

6 .0146

7 .0044

8 .0011

9 .0003
10 .0001

11 .0000

.3504
 P a g e | 275

d) P(2 < x < 5  = 4.2):

P(x=3) + P(x=4) + P(x=5) =

x Prob.

3 .1852
4 .1944
5 .1633

.5429
 P a g e | 276

C 3 1 C1 (10)(1)
5

6 C4 15
5.38 a) P(x = 3 N = 6, n = 4, A = 5) = = .6667

b) P(x < 1 N = 10, n = 3, A = 5):

5 C1 5 C 2 5 C 0 5 C 3 (5)(10) (1)(10)

10 C 3 10 C 3 120 120
P(x = 1) + P(x = 0) = + =

= .4167 + .0833 = .5000

c) P(x > 2  N = 13, n = 5, A = 3):

P(x=2) + P(x=3) Note: only 3 x's in population

3 C 2 10 C 3 3 C 3 10 C 2 (3)(120) (1)( 45)


13 C 5 13 C 5 1287 1287
+ = = .2797 + .0350 = .
3147

5.39 n = 25 p = .20 retired

from Table A.2: P(x = 7) = .111

P(x > 10): P(x = 10) + P(x = 11) + . . . + P(x = 25) = .012 + .004
+ .001 = .017
 P a g e | 277

Expected Value = µ = np = 25(.20) = 5

n = 20 p = .40 mutual funds

P(x = 8) = .180

P(x < 6) = P(x = 0) + P(x = 1) + . . . + P(x = 5) =

.000 + .000 + .003 +.012 + .035 + .075 = .125

P(x = 0) = .000

P(x > 12) = P(x = 12) + P(x = 13) + . . . + P(x = 20) = .035 + .015 + .
005 + .001 = .056

x=8

Expected Number = µ = n p = 20(.40) = 8

5.40  = 3.2 cars 2 hours

a) P(x=3) cars per 1 hour) = ??

The interval has been decreased by ½.

 P a g e | 278

The new  = 1.6 cars 1 hour.

P(x = 3 = 1.6) = (from Table A.3) .1378

b) P(x = 0 cars per ½ hour) = ??

The interval has been decreased by ¼ the original amount.

The new  = 0.8 cars ½ hour.

P(x = 0  = 0.8) = (from Table A.3) .4493

c) P(x > 5  = 1.6) = (from Table A.3)

x Prob.

5 .0176

6 .0047
7 .0011

8 .0002

.0236

Either a rare event occurred or perhaps the long-run average, ,

has changed

(increased).
P a g e | 279
 P a g e | 280

5.41 N = 32 A = 10 n = 12

10 C 3  22 C 9 (120)( 497,420)
32 C12 225,792,840
a) P(x = 3) = = = .2644

10 C 6  22 C 6 (210)(74,613)
32 C12 225,792,840
b) P(x = 6) = = = .0694

10 C 0  22 C12 (1)(646,646)
32 C12 225,792,840
c) P(x = 0) = = = .0029

d) A = 22

22 C 7 10 C 5 22 C8 10 C 4 22 C 9 10 C 3

32 C12 32 C12 32 C12
P(7 < x < 9) = + +

(170,544)( 252) (319,770)( 210) (497,420)(120)

 
225,792,840 225,792,840 225,792,840
=

= .1903 + .2974 + .2644 = .7521

 P a g e | 281

5.42  = 1.4 defects 1 lot If x > 3, buyer rejects If x < 3, buyer

accepts

P(x < 3  = 1.4) = (from Table A.3)

x Prob.

0 .2466

1 .3452

2 .2417

3 .1128

.9463
 P a g e | 282

5.43 a) n = 20 and p = .25

The expected number = µ = np = (20)(.25) = 5.00

b) P(x < 1 n = 20 and p = .25) =

P(x = 1) + P(x = 0) = C1(.25)1(.75)19 +

20 C0(.25)0(.75)20
20

= (20)(.25)(.00423) + (1)(1)(.0032) = .0212 +. 0032 = .0244

Since the probability is so low, the population of your state may

have a lower

percentage of chronic heart conditions than those of other states.

5.44 a) P(x > 7 n = 10 and p = .70) = (from Table A.2):

x Prob.

8 .233

9 .121

10 .028

.382
 P a g e | 283

Expected number = µ = np = 10(.70) = 7

b) n = 15 p = 1/3 Expected number = µ = np = 15(1/3) = 5

P(x=0 n = 15 and p = 1/3) =

15C0(1/3)0(2/3)15 = .0023

c) n = 7 p = .53

P(x = 7 n = 7 and p = .53) = 7C7(.53)7(.47)0 = .0117

Probably the 53% figure is too low for this population since the
probability of

this occurrence is so low (.0117).

 P a g e | 284

5.45 n = 12

a.) P(x = 0 long hours):

p = .20 12 C0(.20)0(.80)12 = .0687

b.) P(x > 6) long hours):

p = .20

Using Table A.2: .016 + .003 + .001 = .020

c) P(x = 5 good financing):

p = .25, C5(.25)5(.75)7 = .1032

12

d.) p = .19 (good plan), expected number = µ = n(p) = 12(.19) =

2.28

5.46 n = 100,000 p = .000014

Worked as a Poisson:  = np = 100,000(.000014) = 1.4

a) P(x = 5):
 P a g e | 285

from Table A.3 = .0111

b) P(x = 0):

from Table A.3 = .2466

c) P(x > 6): (from Table A.3)

x Prob

7 .0005

8 .0001

.0006
 P a g e | 286

5.47 P(x < 3)  n = 8 and p = .60): From Table A.2:

x Prob.

0 .001

1 .008

2 .041

3 .124

.174

17.4% of the time in a sample of eight, three or fewer customers are

walk-ins by

chance. Other reasons for such a low number of walk-ins might be that
she is

retaining more old customers than before or perhaps a new competitor

is

attracting walk-ins away from her.

5.48 n = 25 p = .20

a) P(x = 8 n = 25 and p = .20) = (from Table A.2) .062

b) P(x > 10) n = 25 and p = .20) = (from Table A.2)

 P a g e | 287

x Prob.

11 .004

12 .001

13 .000

.005

c) Since such a result would only occur 0.5% of the time by chance, it
is likely

that the analyst's list was not representative of the entire state of
Idaho or the

20% figure for the Idaho census is not correct.

 P a g e | 288

5.49  = 0.6 flats 2000 miles

P(x = 0  = 0.6) = (from Table A.3) .5488

P(x > 3  = 0.6) = (from Table A.3)

x Prob.

3 .0198

4 .0030

5 .0004

.0232

Assume one trip is independent of the other. Let F = flat tire and NF =
no flat tire

P(NF1  NF2) = P(NF1)  P(NF2)

but P(NF) = .5488

P(NF1  NF2) = (.5488)(.5488) = .3012

5.50 N = 25 n=8

a) P(x = 1 in NY) A=4

 P a g e | 289

4 C1  21 C 7 (4)(116,280)
25 C 8 1,081,575
= = .4300

b) P(x = 4 in top 10) A = 10

C 4 15 C 4 (210(1365)
10

25 C 8 1,081,575
= .2650

c) P(x = 0 in California) A=5

C 0  20 C8 (1)(125,970)
5

C
25 8 1,081,575
= .1165

d) P(x = 3 with M) A=3

C 3  22 C 5 (1)( 26,334)
3

25 C 8 1,081,575
= .0243

5.51 N = 24 n=6 A=8

C 6 16 C 0 (28)(1)
8

24 C 6 134,596
a) P(x = 6) = = .0002
 P a g e | 290

C 0 16 C 6 (1)(8008)
8

24 C 6 134,596
b) P(x = 0) = = .0595

d) A = 16 East Side

C 3 8 C 3 (560)(56)
16

24 C 6 134,596
P(x = 3) = = .2330

5.52 n = 25 p = .20 Expected Value = µ = np = 25(.20) = 5

n  p  q  25(.20)(. 80)
µ = 25(.20) = 5  = = 2

P(x > 12) = From Table A.2: x

Prob
 P a g e | 291

13 .0000

The values for x > 12 are so far away from the expected value that
they are very

unlikely to occur.

P(x = 14) = 25C14(.20)14(.80)11 = .000063 which is very unlikely.

If this value (x = 14) actually occurred, one would doubt the validity of
the

p = .20 figure or one would have experienced a very rare event.

5.53  = 2.4 calls 1 minute

a) P(x = 0 = 2.4) = (from Table A.3) .0907

b) Can handle x < 5 calls Cannot handle x > 5 calls

P(x > 5 = 2.4) = (from Table A.3)

x Prob.

6 .0241

7 .0083
 P a g e | 292

8 .0025

9 .0007

10 .0002

11 .0000

.0358

c) P(x = 3 calls 2 minutes)

The interval has been increased 2 times.

New Lambda:  = 4.8 calls 2 minutes.

from Table A.3: .1517

d) P(x < 1 calls 15 seconds):

The interval has been decreased by ¼.

New Lambda =  = 0.6 calls 15 seconds.

P(x < 1 = 0.6) = (from Table A.3)

P(x = 1) = .3293

P(x = 0) = .5488

.8781
 P a g e | 293

5.54 n = 160 p = .01

Working this problem as a Poisson problem:

a) Expected number = µ = n(p) = 160(.01) = 1.6

b) P(x > 8):

Using Table A.3: x Prob.

8 .0002

9 .0000

.0002

c) P(2 < x < 6):

Using Table A.3: x Prob.

2 .2584

3 .1378

4 .0551
5 .0176

6 .0047

.4736
 P a g e | 294

5.55 p = .005 n = 1,000

 = np = (1,000)(.005) = 5

a) P(x < 4) = P(x = 0) + P(x = 1) + P(x = 2) + P(x = 3) =

.0067 + .0337 + .0842 + .1404 = .265

b) P(x > 10) = P(x = 11) + P(x = 12) + . . . =

.0082 + .0034 + .0013 + .0005 + .0002 = .0136

c) P(x = 0) = .0067
 P a g e | 295

5.56 n=8 p = .36 x = 0 women

8 C0(.36)0(.64)8 = (1)(1)(.0281475) = .0281

It is unlikely that a company would randomly hire 8 physicians from the

U.S. pool

and none of them would be female. If this actually happened, figures

similar to these

might be used as evidence in a lawsuit.

5.57 N = 34

a) n = 5 x=3 A = 13

C 3  21 C 2 (286)( 210)
13

34 C 5 278,256
= .2158

b) n = 8 x<2 A=5

5 C 0  29 C8 5 C1  29 C 7 5 C 2  29 C 6
34 C 8 34 C 8 34 C 8
+ + =
 P a g e | 296

(1)( 4,292,145) (5)(1,560,780) (10)( 475,020)

 
18,156,204 18,156,204 18,156,204
= .2364 + .4298 + .2616 = .
9278

c) n = 5 x=2 A=3

5 C2(3/34)2(31/34)3 = (10)(.0077855)(.7579636) = .0590

 P a g e | 297

5.58 N = 14 n=4

a) P(x = 4 N = 14, n = 4, A = 10 north side)

C 4  4 C 0 (210((1)
10

14 C 4 1001
= .2098

b) P(x = 4 N = 14, n = 4, A = 4 west)

C 4 10 C 0 (1)(1)
4

14 C 4 1001
= .0010

c) P(x = 2 N = 14, n = 4, A = 4 west)

C 2 10 C 2 (6)( 45)

4

14 C 4 1001
= .2697

5.59 a)  = 3.84 1,000

3.840  e 3.84
0!
P(x = 0) = = .0215
 P a g e | 298

b)  = 7.68 2,000

7.686  e 7.68 (205,195.258)(. 000461975)


6! 720
P(x = 6) = = .1317

c)  = 1.6 1,000 and  = 4.8 3,000

from Table A.3:

P(x < 7) = P(x = 0) + P(x = 1) + . . . + P(x = 6) =

.0082 + .0395 + .0948 + .1517 + .1820 + .1747 + .1398 = .7907

 P a g e | 299

5.60 This is a binomial distribution with n = 15 and p = .36.

 = np = 15(.36) = 5.4

15(.36)(.64)
 = = 1.86

The most likely values are near the mean, 5.4. Note from the printout
that the

most probable values are at x = 5 and x = 6 which are near the mean.

5.61 This printout contains the probabilities for various values of x from zero
to eleven from a Poisson distribution with  = 2.78. Note that the
highest probabilities are at x = 2 and

x = 3 which are near the mean. The probability is slightly higher at x

= 2 than at x = 3 even though x = 3 is nearer to the mean because of
the “piling up” effect of x = 0.

5.62 This is a binomial distribution with n = 22 and p = .64.

The mean is np = 22(.64) = 14.08 and the standard deviation is:
 P a g e | 300

n  p  q  22(.64)(. 36)
 = = 2.25

The x value with the highest peak on the graph is at x = 14 followed by

x = 15

and x = 13 which are nearest to the mean.

5.63 This is the graph of a Poisson Distribution with  = 1.784. Note the
high

probabilities at x = 1 and x = 2 which are nearest to the mean. Note

also that the

probabilities for values of x > 8 are near to zero because they are so
far away

from the mean or expected value.

 P a g e | 301

Chapter 6
Continuous Distributions

LEARNING OBJECTIVES

The primary objective of Chapter 6 is to help you understand continuous

distributions, thereby enabling you to:

1. Understand concepts of the uniform distribution.

2. Appreciate the importance of the normal distribution.

3. Recognize normal distribution problems and know how to solve such

problems.

4. Decide when to use the normal distribution to approximate binomial

distribution problems and know how to work such problems.

5. Decide when to use the exponential distribution to solve problems in

business and know how to work such problems.

CHAPTER TEACHING STRATEGY

 P a g e | 302

Chapter 5 introduced the students to discrete distributions. This

chapter introduces the students to three continuous distributions: the uniform
distribution, the normal distribution and the exponential distribution. The
normal distribution is probably the most widely known and used distribution.
The text has been prepared with the notion that the student should be able to
work many varied types of normal curve problems. Examples and practice
problems are given wherein the student is asked to solve for virtually any of
the four variables in the z equation. It is very helpful for the student to get
into the habit of constructing a normal curve diagram, with a shaded portion
for the desired area of concern for each problem using the normal
distribution. Many students tend to be more visual learners than auditory
and these diagrams will be of great assistance in problem demonstration and
in problem solution.

This chapter contains a section dealing with the solution of binomial distribution
problems by the normal curve. The correction for continuity is emphasized. In this text,
the correction for continuity is always used whenever a binomial distribution problem is
worked by the normal curve. Since this is often a stumbling block for students to
comprehend, the chapter has included a table (Table 6.4) with rules of thumb as to how to
apply the correction for continuity. It should be emphasized, however, that answers for
this type of problem are still only approximations. For this reason and also in an effort to
link chapters 5 & 6, the student is sometimes asked to work binomial problems both by
methods in this chapter and also by using binomial tables (A.2). This also will allow the
student to observe how good the approximation of the normal curve is to binomial
problems.
The exponential distribution can be taught as a continuous distribution,
which can be used in complement with the Poisson distribution of chapter 5
to solve inter-arrival time problems. The student can see that while the
Poisson distribution is discrete because it describes the probabilities of whole
number possibilities per some interval, the exponential distribution describes
the probabilities associated with times that are continuously distributed.

CHAPTER OUTLINE

6.1 The Uniform Distribution

Determining Probabilities in a Uniform Distribution

 P a g e | 303

Using the Computer to Solve for Uniform Distribution Probabilities

6.2 Normal Distribution

History of the Normal Distribution

Probability Density Function of the Normal Distribution

Standardized Normal Distribution

Solving Normal Curve Problems

Using the Computer to Solve for Normal Distribution Probabilities

6.3 Using the Normal Curve to Approximate Binomial Distribution Problems

Correcting for Continuity

6.4 Exponential Distribution

Probabilities of the Exponential Distribution

Using the Computer to Determine Exponential Distribution

Probabilities

KEY TERMS

Correction for Continuity Standardized Normal Distribution

Exponential Distribution Uniform Distribution

Normal Distribution z Distribution

Rectangular Distribution z Score

 P a g e | 304

SOLUTIONS TO PROBLEMS IN CHAPTER 6

6.1 a = 200 b = 240

1 1 1
 
b  a 240  200 40
a) f(x) = = .025

a  b 200  240

2 2
b)  = = 220

b  a 240  200 40
 
12 12 12
 = = 11.547

240  230 10

240  200 40
c) P(x > 230) = = .250

220  205 15

240  200 40
d) P(205 < x < 220) = = .375

225  200 25

240  200 40
e) P(x < 225) = = .625
 P a g e | 305

6.2 a = 8 b = 21

1 1 1
 
b  a 21  8 13
a) f(x) = = .0769

a  b 8  21 29
 
2 2 2
b)  = = 14.5

b  a 21  8 13
 
12 12 12
 = = 3.7528

17  10 7

21  8 13
c) P(10 < x < 17) = = .5385

d) P(x > 22) = .0000

e) P(x > 7) = 1.0000

 P a g e | 306

6.3 a = 2.80 b = 3.14

a  b 2.80  3.14

2 2
 = = 2.97

b  a 3.14  2.80

12 12
 = = 0.098

3.10  3.00
3.14  2.80
P(3.00 < x < 3.10) = = 0.2941

6.4 a = 11.97 b = 12.03

1 1

b  a 12.03  11.97
Height = = 16.667

12.03  12.01
12.03  11.97
P(x > 12.01) = = .3333

12.01  11.98
12.03  11.97
P(11.98 < x < 12.01) = = .5000
 P a g e | 307

6.5 µ = 2100 a = 400 b = 3800

b  a 3800  400

12 12
 = = 981.5

1 3800  400

ba 12
Height = = .000294

3800  3000 800


3800  400 3400
P(x > 3000) = = .2353

P(x > 4000) = .0000

1500  700 800


3800  400 3400
P(700 < x < 1500) = = .2353

6.6 a) P(z > 1.96):

Table A.5 value for z = 1.96: .4750

P(z > 1.96) = .5000 - .4750 = .0250

 P a g e | 308

b) P (z < 0.73):

Table A.5 value for z = 0.73: .2673

P(z < 0.73) = .5000 + .2673 = .7673

c) P(-1.46 < z < 2.84):

Table A.5 value for z = 2.84: .4977

Table A.5 value for z = 1.46: .4279

P(1.46 < z < 2.84) = .4977 + 4279 = .9256

d) P(-2.67 < z < 1.08):

Table A.5 value for z = -2.67: .4962

Table A.5 value for z = 1.08: .3599

P(-2.67 < z < 1.08) = .4962 + .3599 = .8561

e) P (-2.05 < z < -.87):

Table A.5 value for z = -2.05: .4798

 P a g e | 309

Table A.5 value for z = -0.87: .3078

P(-2.05 < z < -.87) = .4798 - .3078 = .1720

 P a g e | 310

6.7 a) P(x < 635 µ = 604,  = 56.8):

x   635  604

 56.8
z = = 0.55

Table A.5 value for z = 0.55: .2088

P(x < 635) = .2088 + .5000 = .7088

b) P(x < 20 µ = 48,  = 12):

x   20  48

 12
z = = -2.33

Table A.5 value for z = -2.33: .4901

P(x < 20) = .5000 - .4901 = .0099

c) P(100 < x < 150 µ = 111,  = 33.8):

x   150  111

 33.8
z = = 1.15

Table A.5 value for z = 1.15: .3749

 P a g e | 311

x   100  111

 33.8
z = = -0.33

Table A.5 value for z = -0.33: .1293

P(100 < x < 150) = .3749 + .1293 = .5042

d) P(250 < x < 255 µ = 264,  = 10.9):

x   250  264

 10.9
z = = -1.28

Table A.5 value for z = -1.28: .3997

x   255  264

 10.9
z = = -0.83

Table A.5 value for z = -0.83: .2967

P(250 < x < 255) = .3997 - .2967 = .1030

e) P(x > 35 µ = 37,  = 4.35):

x   35  37

 4.35
z = = -0.46
 P a g e | 312

Table A.5 value for z = -0.46: .1772

P(x > 35) = .1772 + .5000 = .6772

f) P(x > 170 µ = 156,  = 11.4):

x   170  156

 11.4
z = = 1.23

Table A.5 value for z = 1.23: .3907

P(x > 170) = .5000 - .3907 = .1093

6.8 µ = 22  =4

a) P(x > 17):

x   17  22

 4
z = = -1.25

area between x = 17 and µ = 22 from table A.5 is .3944

 P a g e | 313

P(x > 17) = .3944 + .5000 = .8944

b) P(x < 13):

x   13  22

 4
z = = -2.25

from table A.5, area = .4878

P(x < 13) = .5000 - .4878 = .0122

 P a g e | 314

c) P(25 < x < 31):

x   31  22

 4
z = = 2.25

from table A.5, area = .4878

x   25  22

 4
z = = 0.75

from table A.5, area = .2734

P(25 < x < 31) = .4878 - .2734 = .2144

6.9 µ = 60  = 11.35

a) P(x > 85):

x   85  60

 11.35
z = = 2.20

from Table A.5, the value for z = 2.20 is .4861

 P a g e | 315

P(x > 85) = .5000 - .4861 = .0139

b) P(45 < x < 70):

x   45  60

 11.35
z = = -1.32

x   70  60

 11.35
z = = 0.88

from Table A.5, the value for z = -1.32 is .4066

and for z = 0.88 is .3106

P(45 < x < 70) = .4066 + .3106 = .7172

 P a g e | 316

c) P(65 < x < 75):

x   65  60

 11.35
z = = 0.44

x   75  60

 11.35
z = = 1.32

from Table A.5, the value for z = 0.44 is .1700

from Table A.5, the value for z = 1.32 is .4066

P(65 < x < 75) = .4066 - .1700 = .2366

d) P(x < 40):

x   40  60

 11.35
z = = -1.76

from Table A.5, the value for z = -1.76 is .4608

P(x < 40) = .5000 - .4608 = .0392

6.10 µ = $1332  = $725

 P a g e | 317

a) P(x > $2000):

x   2000  1332

 725
z = = 0.92

from Table A.5, the z = 0.92 yields: .3212

P(x > $2000) = .5000 - .3212 = .1788

b) P(owes money) = P(x < 0):

x   0  1332

 725
z = = -1.84

from Table A.5, the z = -1.84 yields: .4671

P(x < 0) = .5000 - .4671 = .0329

c) P($100 < x < $700):

x   100  1332

 725
z = = -1.70

from Table A.5, the z = -1.70 yields: .4554

 P a g e | 318

x   700  1332

 725
z = = -0.87

from Table A.5, the z = -0.87 yields: .3078

P($100 < x < $700) = .4554 - .3078 = .1476

6.11  = $30,000  = $9,000

a) P($15,000 < x < $45,000):

x   45,000  30,000

 9,000
z = = 1.67

From Table A.5, z = 1.67 yields: .4525

x   15,000  30,000

 9,000
z = = -1.67

From Table A.5, z = -1.67 yields: .4525

P($15,000 < x < $45,000) = .4525 + .4525 = .9050

 P a g e | 319

b) P(x > $50,000):

x   50,000  30,000

 9,000
z = = 2.22

From Table A.5, z = 2.22 yields: .4868

P(x > $50,000) = .5000 - .4868 = .0132

 P a g e | 320

c) P($5,000 < x < $20,000):

x   5,000  30,000

 9,000
z = = -2.78

From Table A.5, z = -2.78 yields: .4973

x   20,000  30,000

 9,000
z = = -1.11

From Table A.5, z = -1.11 yields .3665

P($5,000 < x < $20,000) = .4973 - .3665 = .1308

 P a g e | 321

d) Since 90.82% of the values are greater than x = $7,000, x = $7,000

is in the

lower half of the distribution and .9082 - .5000 = .4082 lie between
x and µ.

From Table A.5, z = -1.33 is associated with an area of .4082.

x

Solving for : z =

7,000  30,000

-1.33 =

 = 17,293.23

e)  = $9,000. If 79.95% of the costs are less than $33,000, x =

$33,000 is in

the upper half of the distribution and .7995 - .5000 = .2995 of the
values lie

between $33,000 and the mean.

From Table A.5, an area of .2995 is associated with z = 0.84

 P a g e | 322

x

Solving for µ: z =

33,000  
9,000
0.84 =

µ = $25,440

6.12 µ = 200,  = 47 Determine x

a) 60% of the values are greater than x:

Since 50% of the values are greater than the mean, µ = 200, 10%
or .1000 lie

between x and the mean. From Table A.5, the z value associated
with an area

of .1000 is z = -0.25. The z value is negative since x is below the

mean.

Substituting z = -0.25, µ = 200, and  = 47 into the formula and

solving for x:

x

z =

x  200
47
-0.25 =
 P a g e | 323

x = 188.25

b) x is less than 17% of the values.

Since x is only less than 17% of the values, 33% (.5000- .1700) or .
3300 lie

between x and the mean. Table A.5 yields a z value of 0.95 for an
area of

.3300. Using this z = 0.95, µ = 200, and  = 47, x can be solved

for:

x

z =

x  200
47
0.95 =

x = 244.65

c) 22% of the values are less than x.

Since 22% of the values lie below x, 28% lie between x and the
mean

(.5000 - .2200). Table A.5 yields a z of -0.77 for an area of .2800.

Using the z

value of -0.77, µ = 200, and  = 47, x can be solved for:

 P a g e | 324

x

z =

x  200
47
-0.77 =

x = 163.81

d) x is greater than 55% of the values.

Since x is greater than 55% of the values, 5% (.0500) lie between x

and the

mean. From Table A.5, a z value of 0.13 is associated with an area

of .05.

Using z = 0.13, µ = 200, and  = 47, x can be solved for:

x

z =

x  200
47
0.13 =

x = 206.11
 P a g e | 325

6.13  = 625. If 73.89% of the values are greater than 1700, then
23.89% or .2389

lie between 1700 and the mean, µ. The z value associated with .
2389 is -0.64

since the 1700 is below the mean.

Using z = -0.64, x = 1700, and  = 625, µ can be solved for:

x

z =

1700  
625
-0.64 =

µ = 2100

µ = 2258 and  = 625. Since 31.56% are greater than x, 18.44% or

.1844

(.5000 - .3156) lie between x and µ = 2258. From Table A.5, a z

value of 0.48

is associated with .1844 area under the normal curve.

Using µ = 2258,  = 625, and z = 0.48, x can be solved for:

x

z =
 P a g e | 326

x  2258
625
0.48 =

x = 2558

6.14 µ = 22  = ??

Since 72.4% of the values are greater than 18.5, then 22.4% lie
between 18.5 and µ. x = 18.5 is below the mean. From table A.5, z =
- 0.59.

18.5  22

-0.59 =

-0.59 = -3.5

3.5
0.59
 = = 5.932

6.15 P(x < 20) = .2900

x is less than µ because of the percentage. Between x and µ is .5000

- .2900 =
 P a g e | 327

.2100 of the area. The z score associated with this area is -0.55.
Solving for µ:

x

z =

20  
4
-0.55 =

 = 22.20

6.16 µ = 9.7 Since 22.45% are greater than 11.6, x = 11.6 is in the
upper half of the distribution and .2755 (.5000 - .2245) lie between x
and the mean. Table A.5 yields a z = 0.76 for an area of .2755.

Solving for :

x

z =

11.6  9.7

0.76 =

 = 2.5
 P a g e | 328

6.17 a) P(x < 16 n = 30 and p = .70)

µ = np = 30(.70) = 21

n  p  q  30(.70)(.30)
 = = 2.51

P(x < 16.5 µ = 21 and  = 2.51)

b) P(10 < x < 20 n = 25 and p = .50)

µ = np = 25(.50) = 12.5

n  p  q  25(.50)(. 50)
 = = 2.5

P(10.5 < x < 20.5 µ = 12.5 and  = 2.5)

c) P(x = 22 n = 40 and p = .60)

µ = np = 40(.60) = 24
 P a g e | 329

n  p  q  40(.60)(. 40)
 = = 3.10

P(21.5 < x < 22.5 µ = 24 and  = 3.10)

d) P(x > 14 n = 16 and p = .45)

µ = np = 16(.45) = 7.2

n  p  q  16(.45)(.55)
 = = 1.99

P(x > 14.5 µ = 7.2 and  = 1.99)

 P a g e | 330

6.18 a) n = 8 and p = .50 µ = np = 8(.50) = 4

n  p  q  8(.50)(. 50)
 = = 1.414

µ ± 3 = 4 ± 3(1.414) = 4 ± 4.242

(-0.242 to 8.242) does not lie between 0 and 8.

Do not use the normal distribution to approximate this problem.

b) n = 18 and p = .80 µ = np = 18(.80) = 14.4

n  p  q  18(.80)(.20)
 = = 1.697

µ ± 3 = 14.4 ± 3(1.697) = 14.4 ± 5.091

(9.309 to 19.491) does not lie between 0 and 18.

Do not use the normal distribution to approximate this problem.

c) n = 12 and p = .30

µ = np = 12(.30) = 3.6

 P a g e | 331

n  p  q  12(.30)(. 70)
 = = 1.587

µ ± 3 = 3.6 ± 3(1.587) = 3.6 ± 4.761

(-1.161 to 8.361) does not lie between 0 and 12.

Do not use the normal distribution to approximate this problem.

d) n = 30 and p = .75 µ = np = 30(.75) = 22.5

µ ± 3 = 22.5 ± 3(2.37) = 22.5 ± 7.11

(15.39 to 29.61) does lie between 0 and 30.

The problem can be approximated by the normal curve.

e) n = 14 and p = .50 µ = np = 14(.50) = 7

n  p  q  14(.50)(. 50)
 = = 1.87

µ ± 3 = 7 ± 3(1.87) = 7 ± 5.61

(1.39 to 12.61) does lie between 0 and 14.

The problem can be approximated by the normal curve.

6.19 a) P(x = 8 n = 25 and p = .40) µ = np = 25(.40) = 10

 P a g e | 332

n  p  q  25(.40)(. 60)
 = = 2.449

µ ± 3 = 10 ± 3(2.449) = 10 ± 7.347

(2.653 to 17.347) lies between 0 and 25.

Approximation by the normal curve is sufficient.

P(7.5 < x < 8.5 µ = 10 and  = 2.449):

7.5  10
2.449
z = = -1.02

From Table A.5, area = .3461

8.5  10
2.449
z = = -0.61

From Table A.5, area = .2291

P(7.5 < x < 8.5) = .3461 - .2291 = .1170

From Table A.2 (binomial tables) = .120

b) P(x > 13 n = 20 and p = .60) µ = np = 20(.60) = 12

 P a g e | 333

n  p  q  20(.60)(. 40)
 = = 2.19

µ ± 3 = 12 ± 3(2.19) = 12 ± 6.57

(5.43 to 18.57) lies between 0 and 20.

Approximation by the normal curve is sufficient.

P(x > 12.5 µ = 12 and  = 2.19):

x   12.5  12

 2.19
z = = 0.23

From Table A.5, area = .0910

P(x > 12.5) = .5000 -.0910 = .4090

From Table A.2 (binomial tables) = .415

c) P(x = 7 n = 15 and p = .50) µ = np = 15(.50) = 7.5

n  p  q  15(.50)(.50)
 = = 1.9365

µ ± 3 = 7.5 ± 3(1.9365) = 7.5 ± 5.81

(1.69 to 13.31) lies between 0 and 15.

 P a g e | 334

Approximation by the normal curve is sufficient.

P(6.5 < x < 7.5 µ = 7.5 and  = 1.9365):

x   6.5  7.5

 1.9365
z = = -0.52

From Table A.5, area = .1985

From Table A.2 (binomial tables) = .196

d) P(x < 3 n = 10 and p =.70): µ = np = 10(.70) = 7

n  p  q  10(.70)(.30)
 =

µ ± 3 = 7 ± 3(1.449) = 7 ± 4.347

(2.653 to 11.347) does not lie between 0 and 10.

The normal curve is not a good approximation to this problem.

6.20 P(x < 40 n = 120 and p = .37): µ = np = 120(.37) =

44.4
 P a g e | 335

n  p  q  120(.37)(. 63)
 = = 5.29

µ + 3 = 28.53 to 60.27 does lie between 0 and 120.

It is okay to use the normal distribution

to approximate this problem

Correcting for continuity: x = 39.5

39.5  44.4
5.29
z = = -0.93

from Table A.5, the area of z = -0.93 is .3238

P(x < 40) = .5000 - .3238 = .1762

6.21 n = 70, p = .59 P(x < 35):

Converting to the normal dist.:

n  p  q  70(.59)(.41)
µ = n(p) = 70(.59) = 41.3 and  = = 4.115

Test for normalcy:

0 < µ + 3 < n, 0 < 41.3 + 3(4.115) < 70

0 < 28.955 to 53.645 < 70, passes the test

 P a g e | 336

correction for continuity, use x = 34.5

34.5  41.3
4.115
z = = -1.65

from table A.5, area = .4505

P(x < 35) = .5000 - .4505 = .0495

6.22 For parts a) and b), n = 300 p = .53

µ = 300(.53) = 159

n  p  q  300(.53)(.47)
 = = 8.645

Test: µ + 3 = 159 + 3(8.645) = 133.065 to 184.935

which lies between 0 and 300. It is okay to use the normal distribution
as an approximation on parts a) and b).

a) P(x > 175 transmission)

correcting for continuity: x = 175.5

 P a g e | 337

175.5  159
8.645
z = = 1.91

from A.5, the area for z = 1.91 is .4719

P(x > 175) = .5000 - .4719 = .0281

b) P(165 < x < 170)

correcting for continuity: x = 164.5; x = 170.5

170.5  159 164.5  159

8.645 8.645
z = = 1.33 and z = = 0.64

from A.5, the area for z = 1.33 is .4082

the area for z = 0.64 is .2389

P(165 < x < 170) = .4082 - .2389 = .1693

For parts c) and d): n = 300 p = .60

n  p  q  300(.60)(.40)
 = 300(.60) = 180  = =
8.485
 P a g e | 338

Test: µ + 3 = 180 + 3(8.485) = 180 + 25.455

154.545 to 205.455 lies between 0 and 300

It is okay to use the normal distribution

to approximate c) and d)

c) P(155 < x < 170 personnel):

correcting for continuity: x = 154.5; x = 170.5

170.5  180 154.5  180

8.485 8.485
z = = -1.12 and z = = -3.01

from A.5, the area for z = -1.12 is .3686

the area for z = -3.01 is .4987

P(155 < x < 170) = .4987 - .3686 = .1301

d) P(x < 200 personnel):

correcting for continuity: x = 199.5

199.5  180
8.485
z = = 2.30

from A.5, the area for z = 2.30 is .4893

 P a g e | 339

P(x < 200) = .5000 + .4893 = .9893

6.23 p = .25 n = 130

Conversion to normal dist.: µ = n(p) = 130(.25) = 32.5

n  p  q  130(.25)(.75)
 = = 4.94

a) P(x > 36): Correct for continuity: x = 36.5

36.5  32.5
4.94
z = = 0.81

from table A.5, area = .2910

P(x > 20) = .5000 - .2910 = .2090

b) P(26 < x < 35): Correct for continuity: 25.5 to 35.5

25.5  32.5 35.5  32.5

4.94 4.94
z = = -1.42 and z = = 0.61

from table A.5, area for z = -1.42 is .4222

area for z = 0.61 is .2291

P(26 < x < 35) = .4222 + .2291 = .6513

 P a g e | 340

c) P(x < 20): correct for continuity: x = 19.5

19.5  32.5
4.94
z = = -2.63

from table A.5, area for z = -2.63 is .4957

P(x < 20) = .5000 - .4957 = .0043

d) P(x = 30): correct for continuity: 29.5 to 30.5

29.5  32.5 30.5  32.5

4.94 4.94
z = = -0.61 and z = = -0.40

from table A.5, area for -0.61 = .2291

area for -0.40 = .1554

P(x = 30) = .2291 - .1554 = .0737

6.24 n = 95

a) P(44 < x < 52) agree with direct investments, p = .52

By the normal distribution: µ = n(p) = 95(.52) = 49.4

 P a g e | 341

n  p  q  95(.52)(. 48)
 = = 4.87

test: µ + 3 = 49.4 + 3(4.87) = 49.4 + 14.61

0 < 34.79 to 64.01 < 95 test passed

43.5  49.4
4.87
z = = -1.21

from table A.5, area = .3869

52.5  49.4
4.87
z = = 0.64

from table A.5, area = .2389

P(44 < x < 52) = .3869 + .2389 = .6258

b) P(x > 56):

correcting for continuity, x = 56.5

56.5  49.4
4.87
z = = 1.46
 P a g e | 342

from table A.5, area = .4279

P(x > 56) = .5000 - .4279 = .0721

c) Joint Venture:

p = .70, n = 95

By the normal dist.: µ = n(p) = 95(.70) = 66.5

n  p  q  95(.70)(. 30)
 = = 4.47

test for normalcy: 66.5 + 3(4.47) = 66.5 + 13.41

0 < 53.09 to 79.91 < 95 test passed

P(x < 60):

correcting for continuity: x = 59.5

59.5  66.5
4.47
z = = -1.57

from table A.5, area = .4418

 P a g e | 343

P(x < 60) = .5000 - .4418 = .0582

d) P(55 < x < 62):

correcting for continuity: 54.5 to 62.5

54.5  66.5
4.47
z = = -2.68

from table A.5, area = .4963

62.5  66.5
4.47
z = = -0.89

from table A.5, area = .3133

P(55 < x < 62) = .4963 - .3133 = .1830

 P a g e | 344

6.25 a)  = 0.1

x0 y

0 .1000

1 .0905
2 .0819

3 .0741

4 .0670

5 .0607

6 .0549

7 .0497

8 .0449

9 .0407

10 .0368

b)  = 0.3

x0 y

0 .3000
 P a g e | 345

1 .2222

2 .1646

3 .1220

4 .0904

5 .0669

6 .0496

7 .0367

8 .0272

9 .0202

c)  = 0.8

x0 y

0 .8000

1 .3595

2 .1615

3 .0726
 P a g e | 346

4 .0326

5 .0147

6 .0066

7 .0030

8 .0013

9 .0006

d)  = 3.0

x0 y

0 3.0000

1 .1494

2 .0074

3 .0004
 P a g e | 347

4 .0000

5 .0000
 P a g e | 348

6.26 a)  = 3.25

1 1

 3.25
µ= = 0.31

1 1

 3.25
 = = 0.31

b)  = 0.7

1 1

 .007
µ = = 1.43

1 1

 .007
 = = 1.43

c)  = 1.1

1 1

 1.1
µ = = 0.91

1 1

 1.1
 = = 0.91
 P a g e | 349

d)  = 6.0

1 1

 6
 = = 0.17

1 1

 6
 = = 0.17
 P a g e | 350

6.27 a) P(x > 5  = 1.35) =

for x0 = 5: P(x) = e-x = e-1.35(5) = e-6.75 = .0012

b) P(x < 3  = 0.68) = 1 - P(x < 3  = .68) =

for x0 = 3: 1 – e-x = 1 – e-0.68(3) = 1 – e –2.04

= 1 - .1300 = .
8700

c) P(x > 4  = 1.7) =

for x0 = 4: P(x) = e-x = e-1.7(4) = e-6.8 = .0011

d) P(x < 6  = 0.80) = 1 - P(x > 6  = 0.80) =

for x0 = 6: P(x) = 1 – e-x = 1 – e-0.80(6) = 1 – e-4.8 = 1 - .0082

= .9918

6.28 µ = 23 sec.

1

 = = .0435 per second
 P a g e | 351

a) P(x > 1 min = .0435/sec.)

Change  to minutes:  = .0435(60) = 2.61 min

P(x > 1 min = 2.61/min) =

for x0 = 1: P(x) = e-x = e-2.61(1) = .0735

b)  = .0435/sec

Change  to minutes:  = (.0435)(60) = 2.61 min

P(x > 3 
 = 2.61/min) =

for x0 = 3: P(x) = e-x = e-2.61(3) = e-7.83 = .0004

6.29  = 2.44/min.

a) P(x > 10 min  = 2.44/min) =

Let x0 = 10, e-x = e-2.44(10) = e-24.4 = .0000

 P a g e | 352

b) P(x > 5 min  = 2.44/min) =

Let x0 = 5, e-x = e-2.44(5) = e-12.20 = .0000

c) P(x > 1 min  = 2.44/min) =

Let x0 = 1, e-x = e-2.44(1) = e-2.44 = .0872

1 1

 2.44
d) Expected time = µ = min. = .41 min = 24.6 sec.

6.30  = 1.12 planes/hr.

1 1

 1.12
a) µ = = .89 hr. = 53.4 min.

b) P(x > 2 hrs  = 1.12 planes/hr.) =

Let x0 = 2, e-x = e-1.12(2) = e-2.24 = .1065

c) P(x < 10 min  = 1.12/hr.) = 1 - P(x > 10 min = 1.12/hr.)

Change  to 1.12/60 min. = .01867/min.

 P a g e | 353

1 - P(x > 10 min  = .01867/min) =

Let x0 = 10, 1 – e-x = 1 – e-.01867(10) = 1 – e-.1867 = 1 - .8297 = .

1703
 P a g e | 354

6.31  = 3.39/ 1000 passengers

1 1

 3.39
µ= = 0.295

(0.295)(1,000) = 295

P(x > 500):

Let x0 = 500/1,000 passengers = .5

e-x = e-3.39(.5) = e-1.695 = .1836

P(x < 200):

Let x0 = 200/1,000 passengers = .2

e-x = e-3.39(.2) = e-.678 = .5076

P(x < 200) = 1 - .5076 = .4924

 P a g e | 355

6.32 µ = 20 years

1
20
 = = .05/year

x0 Prob(x > x0)=e-x

1 .9512

2 .9048

3 .8607

If the foundation is guaranteed for 2 years, based on past history,

90.48% of the

foundations will last at least 2 years without major repair and only
9.52% will

require a major repair before 2 years.

 P a g e | 356

6.33  = 2/month

1 1

 2
Average number of time between rain = µ = month = 15
days

 = µ = 15 days

P(x < 2 days  = 2/month):

2
30
Change  to days:  = = .067/day

P(x < 2 days  = .067/day) =

1 – P(x > 2 days  = .067/day)

let x0 = 2, 1 – e-x = 1 – e-.067(2) = 1 – .8746 = .1254

6.34 a = 6 b = 14

1 1 1
 
b  a 14  6 8
f(x) = = .125

a  b 6  14

2 2
 = = 10
 P a g e | 357

b  a 14  6 8
 
12 12 12
 = = 2.309

14  11 3

14  6 8
P(x > 11) = = .375

12  7 5

14  6 8
P(7 < x < 12) = = .625

6.35 a) P(x < 21 µ = 25 and  = 4):

x   21  25

 4
z = = -1.00

From Table A.5, area = .3413

P(x < 21) = .5000 -.3413 = .1587

 P a g e | 358

b) P(x > 77 µ = 50 and  = 9):

x   77  50

 9
z = = 3.00

From Table A.5, area = .4987

P(x > 77) = .5000 -.4987 = .0013

c) P(x > 47 µ = 50 and  = 6):

x   47  50

 6
z = = -0.50

From Table A.5, area = .1915

P(x > 47) = .5000 + .1915 = .6915

d) P(13 < x < 29 µ = 23 and  = 4):

x   13  23

 4
z = = -2.50

From Table A.5, area = .4938

 P a g e | 359

x   29  23

 4
z = = 1.50

From Table A.5, area = .4332

P(13 < x < 29) = .4938 + 4332 = .9270

e) P(x > 105 µ = 90 and  = 2.86):

x   105  90

 2.86
z = = 5.24

From Table A.5, area = .5000

P(x > 105) = .5000 - .5000 = .0000

 P a g e | 360

6.36 a) P(x = 12 n = 25 and p = .60):

µ = np = 25(.60) = 15

n  p  q  25(.60)(. 40)
= = 2.45

µ ± 3 = 15 ± 3(2.45) = 15 ± 7.35

(7.65 to 22.35) lies between 0 and 25.

The normal curve approximation is sufficient.

P(11.5 < x < 12.5 µ = 15 and  = 2.45):

x   11.5  15

 2.45
z= = -1.43 From Table A.5, area = .4236

x   12.5  15

 2.45
z= = -1.02 From Table A.5, area = .3461

P(11.5 < x < 12.5) = .4236 - .3461 = .0775

From Table A.2, P(x = 12) = .076

b) P(x > 5 n = 15 and p = .50):

 P a g e | 361

µ = np = 15(.50) = 7.5

n  p  q  15(.50)(.50)
= = 1.94

µ ± 3 = 7.5 ± 3(1.94) = 7.5 ± 5.82

(1.68 to 13.32) lies between 0 and 15.

The normal curve approximation is sufficient.

P(x > 5.5 µ = 7.5 and = l.94)

5.5  7.5
1.94
z = = -1.03

From Table A.5, area = .3485

P(x > 5.5) = .5000 + .3485 = .8485

Using table A.2, P(x > 5) = .849

 P a g e | 362

c) P(x < 3 n = 10 and p = .50):

µ = np = 10(.50) = 5

n  p  q  10(.50)(. 50)
= = 1.58

µ ± 3 = 5 ± 3(1.58) = 5 ± 4.74

(0.26 to 9.74) lies between 0 and 10.

The normal curve approximation is sufficient.

P(x < 3.5 µ = 5 and  = l.58):

3 .5  5
1.58
z = = -0.95

From Table A.5, area = .3289

P(x < 3.5) = .5000 - .3289 = .1711

d) P(x > 8 n = 15 and p = .40):

µ = np = 15(.40) = 6
 P a g e | 363

n  p  q  15(.40)(.60)
= = 1.90

 ± 3 = 6 ± 3(1.90) = 6 ± 5.7

(0.3 to 11.7) lies between 0 and 15.

The normal curve approximation is sufficient.

P(x > 7.5 µ = 6 and  = l.9):

7.5  6
1.9
z = = 0.79

From Table A.5, area = .2852

P(x > 7.5) = .5000 - .2852 = .2148

 P a g e | 364

6.37 a) P(x > 3  = 1.3):

let x0 = 3

P(x > 3  = 1.3) = e-x = e-1.3(3) = e-3.9 = .0202

b) P(x < 2  = 2.0):

Let x0 = 2

P(x < 2  = 2.0) = 1 - P(x > 2  = 2.0) =

1 – e-x = 1 – e-2(2) = 1 – e-4 = 1 - .0183 = .9817

c) P(1 < x < 3  = 1.65):

P(x > 1  = 1.65):

Let x0 = 1

e-x = e-1.65(1) = e-1.65 = .1920

P(x > 3 = 1.65):

Let x0 = 3
 P a g e | 365

e-x = e-1.65(3) = e-4.95 = .0071

P(1 < x < 3) = P(x > 1) - P(x > 3) = .1920 - .0071 = .1849

d) P(x > 2 = 0.405):

Let x0 = 2

e-x = e-(.405)(2) = e-.81 = .4449

 P a g e | 366

6.38 µ = 43.4

12% more than 48. x = 48

Area between x and µ is .50 - .12 = .38

z associated with an area of .3800 is z = 1.175

Solving for :

x

z =

48  43.4

1.175 =

4.6
1.175
 = = 3.915
 P a g e | 367

6.39 p = 1/5 = .20 n = 150

P(x > 50):

µ = 150(.20) = 30

150(.20)(.80)
 = = 4.899

50.5  30
4.899
z = = 4.18

Area associated with z = 4.18 is .5000

P(x > 50) = .5000 - .5000 = .0000

 P a g e | 368

6.40  = 1 customer/20 minutes

µ = 1/  = 1

a) 1 hour interval

x0 = 3 because 1 hour = 3(20 minute intervals)

P(x > x0) = e-x = e-1(3) = e-3 = .0498

b) 10 to 30 minutes

x0 = .5, x0 = 1.5

P(x > .5) = e-x = e-1(.5) = e-.5 = .6065

P(x > 1.5) = e-x = e-1(1.5) = e-1.5 = .2231

P(10 to 30 minutes) = .6065 - .2231 = .3834

c) less than 5 minutes

x0 = 5/20 = .25

P(x > .25) = e-x = e-1(.25) = e-.25 = .7788

 P a g e | 369

P(x < .25) = 1 - .7788 = .2212

6.41 µ = 90.28  = 8.53

P(x < 80):

80  90.28
8.53
z = = -1.21

from Table A.5, area for z = -1.21 is .3869

P(x < 80) = .5000 - .3869 = .1131

P(x > 95):

95  90.28
8.53
z = = 0.55

from Table A.5, area for z = 0.55 is .2088

P(x > 95) = .5000 - .2088 = .2912

P(83 < x < 87):

 P a g e | 370

83  90.28
8.53
z = = -0.85

87  90.28
8.53
z = = -0.38

from Table A.5, area for z = -0.85 is .3023

area for z = -0.38 is .1480

P(83 < x < 87) = .3023 - .1480 = .1543

6.42  = 83

Since only 3% = .0300 of the values are greater than 2,655(million), x

= 2655
lies in the upper tail of the distribution. .5000 - .0300 = .4700 of the
values lie

between 2655 and the mean.

Table A.5 yields a z = 1.88 for an area of .4700.

Using z = 1.88, x = 2655,  = 83, µ can be solved for.

x

z =
 P a g e | 371

2655  
83
1.88 =

 = 2498.96 million
 P a g e | 372

6.43 a = 18 b = 65

50  25 25

65  18 47
P(25 < x < 50) = = .5319

a  b 65  18

2 2
 = = 41.5

1 1 1
 
b  a 65  18 47
f(x) = = .0213

6.44  = 1.8 per 15 seconds

1 1

 1 .8
a)  = = .5556(15 sec.) = 8.33 sec.

b) For x0 > 25 sec. use x0 = 25/15 = 1.67

P(x0 > 25 sec.) = e-1.6667(1.8) = .0498

c) x0 < 5 sec. = 1/3

P(x0 < 5 sec.) = 1 - e-1/3(1.8) = 1 - .5488 = .4512

 P a g e | 373

d) P(x0 > 1 min.):

x0 = 1 min. = 60/15 = 4

P(x0 > 1 min.) = e-4(1.8) = .0007

6.45 µ = 951  = 96

a) P(x > 1000):

x   1000  951

 96
z = = 0.51

from Table A.5, the area for z = 0.51 is .1950

P(x > 1000) = .5000 - .1950 = .3050

 P a g e | 374

b) P(900 < x < 1100):

x   900  951

 96
z = = -0.53

x   1100  951

 96
z = = 1.55

from Table A.5, the area for z = -0.53 is .2019

the area for z = 1.55 is .4394

P(900 < x < 1100) = .2019 + .4394 = .6413

c) P(825 < x < 925):

x   825  951

 96
z = = -1.31

x   925  951

 96
z = = -0.27

from Table A.5, the area for z = -1.31 is .4049

the area for z = -0.27 is .1064

P(825 < x < 925) = .4049 - .1064 = .2985

 P a g e | 375

d) P(x < 700):

x   700  951

 96
z = = -2.61

from Table A.5, the area for z = -2.61 is .4955

P(x < 700) = .5000 - .4955 = .0045

 P a g e | 376

6.46 n = 60 p = .24

µ = np = 60(.24) = 14.4

n  p  q  60(.24)(. 76)
= = 3.308

test: µ + 3 = 14.4 + 3(3.308) = 14.4 + 9.924 = 4.476 and 24.324

Since 4.476 to 24.324 lies between 0 and 60, the normal distribution
can be used

to approximate this problem.

P(x > 17):

correcting for continuity: x = 16.5

x   16.5  14.4

 3.308
z = = 0.63

from Table A.5, the area for z = 0.63 is .2357

P(x > 17) = .5000 - .2357 = .2643

P(x > 22):

 P a g e | 377

correcting for continuity: x = 22.5

x   22.5  14.4

 3.308
z = = 2.45

from Table A.5, the area for z = 2.45 is .4929

P(x > 22) = .5000 - .4929 = .0071

P(8 < x < 12):

correcting for continuity: x = 7.5 and x = 12.5

x   12.5  14.4 x   7.5  14.4

 
 3.308  3.308
z = = -0.57 z = = -2.09

from Table A.5, the area for z = -0.57 is .2157

the area for z = -2.09 is .4817

P(8 < x < 12) = .4817 - .2157 = .2660

6.47 µ = 45,970  = 4,246

a) P(x > 50,000):

 P a g e | 378

x   50,000  45,970

 4,246
z = = 0.95

from Table A.5, the area for z = 0.95 is .3289

P(x > 50,000) = .5000 - .3289 = .1711

b) P(x < 40,000):

x   40,000  45,970

 4,246
z = = -1.41

from Table A.5, the area for z = -1.41 is .4207

P(x < 40,000) = .5000 - .4207 = .0793

c) P(x > 35,000):

x   35,000  45,970

 4,246
z = = -2.58

from Table A.5, the area for z = -2.58 is .4951

P(x > 35,000) = .5000 + .4951 = .9951

d) P(39,000 < x < 47,000):

 P a g e | 379

x   39,000  45,970

 4,246
z = = -1.64

x   47,000  45,970

 4,246
z = = 0.24

from Table A.5, the area for z = -1.64 is .4495

the area for z = 0.24 is .0948

P(39,000 < x < 47,000) = .4495 + .0948 = .5443

 P a g e | 380

6.48 µ = 9 minutes

 = 1/µ = .1111/minute = .1111(60)/hour

 = 6.67/hour

P(x > 5 minutes  = .1111/minute) =

1 - P(x > 5 minutes  =.1111/minute):

Let x0 = 5

P(x > 5 minutes  = .1111/minute) =

e-x = e-.1111(5) = e-.5555 = .5738

P(x < 5 minutes) = 1 - P(x > 5 minutes) = 1 - .5738 = .4262

6.49  = 88  = 6.4

a) P(x < 70):

x   70  88

 6.4
z = = -2.81
 P a g e | 381

From Table A.5, area = .4975

P(x < 70) = .5000 - .4975 = .0025

b) P(x > 80):

x   80  88

 6.4
z = = -1.25

From Table A.5, area = .3944

P(x > 80) = .5000 + .3944 = .8944

c) P(90 < x < 100):

x   100  88

 6.4
z = = 1.88

From Table A.5, area = .4699

x   90  88

 6.4
z = = 0.31

From Table A.5, area = .1217

 P a g e | 382

P(90 < x < 100) = .4699 - .1217 = .3482

6.50 n = 200, p = .81

expected number = µ = n(p) = 200(.81) = 162

µ = 162

n  p  q  200(.81)(.19)
 = = 5.548

µ + 3 = 162 + 3(5.548) lie between 0 and 200, the normalcy test is

passed

P(150 < x < 155):

correction for continuity: 150.5 to 154.5

150.5  162
5.548
z = = -2.07

from table A.5, area = .4808

 P a g e | 383

154.5  162
5.548
z = = -1.35

from table A.5, area = .4115

P(150 < x < 155) = .4808 - .4115 = .0693

P(x > 158):

correcting for continuity, x = 158.5

158.5  162
5.548
z = = -0.63

from table A.5, area = .2357

P(x > 158) = .2357 + .5000 = .7357

P(x < 144):

correcting for continuity, x = 143.5

143.5  162
5.548
z = = -3.33
 P a g e | 384

from table A.5, area = .4996

P(x < 144) = .5000 - .4996 = .0004

6.51 n = 150 p = .75

µ = np = 150(.75) = 112.5

n  p  q  150(.75)(.25)
 = = 5.3033

a) P(x < 105):

correcting for continuity: x = 104.5

x   104.5  112.5

 5.3033
z = = -1.51

from Table A.5, the area for z = -1.51 is .4345

P(x < 105) = .5000 - .4345 = .0655

b) P(110 < x < 120):

 P a g e | 385

correcting for continuity: x = 109.5, x = 120.5

109.5  112.5
5.3033
z = = -0.57

120.5  112.5
5.3033
z = = 1.51

from Table A.5, the area for z = -0.57 is .2157

the area for z = 1.51 is .4345

P(110 < x < 120) = .2157 + .4345 = .6502

c) P(x > 95):

correcting for continuity: x = 95.5

95.5  112.5
5.3033
z = = -3.21

from Table A.5, the area for -3.21 is .4993

P(x > 95) = .5000 + .4993 = .9993

 P a g e | 386

ab
2
6.52  = = 2.165

a + b = 2(2.165) = 4.33

b = 4.33 - a

1
ba
Height = = 0.862

1 = 0.862b - 0.862a

Substituting b from above, 1 = 0.862(4.33 - a) - 0.862a

1 = 3.73246 - 0.862a - 0.862a

1 = 3.73246 - 1.724a

1.724a = 2.73246

a = 1.585 and b = 4.33 - 1.585 = 2.745

6.53 µ = 85,200

60% are between 75,600 and 94,800

94,800 –85,200 = 9,600

 P a g e | 387

75,600 – 85,200 = 9,600

The 60% can be split into 30% and 30% because the two x values are
equal distance from the mean.

The z value associated with .3000 area is 0.84

x

z =

94,800  85,200

.84 =

 = 11,428.57

6.54 n = 75 p = .81 prices p = .44 products

a) Expected Value: 1 = np = 75(.81) = 60.75 seeking price

information

n  p  q  75(.81)(. 19)
1 = = 3.3974

b) Expected Value: 2 = np = 75(.44) = 33

 P a g e | 388

n  p  q  75(.44)(. 56)
2 = = 4.2988

Tests: µ + 3 = 60.75 + 3(3.397) = 60.75 + 10.191

50.559 to 70.941 lies between 0 and 75. It is okay to use the

normal

distribution to approximate this problem.

µ + 3 = 33 + 3(4.299) = 33 + 12.897

20.103 to 45.897 lies between 0 and 75. It is okay to use the normal

distribution to approximate this problem.

c) P(x > 67 prices)

correcting for continuity: x = 66.5

66.5  60.75
3.3974
z = = 1.69

from Table A.5, the area for z = 1.69 is .4545

P(x > 67 prices) = .5000 - .4545 = .0455

 P a g e | 389

d) P(x < 23 products):

correcting for continuity: x = 22.5

22.5  33
4.2988
z = = -2.44

from Table A.5, the area for z = -2.44 is .4927

P(x < 23) = .5000 - .4927 = .0073

6.55  = 3 hurricanes 5 months

P(x > 1 month  = 3 hurricanes per 5 months):

Since x and  are for different intervals,

change Lambda =  = 3/ 5 months = 0.6 month.

P(x > month  = 0.6 per month):

Let x0 = 1

P(x > 1) = e-x = e-0.6(1) = e-0.6 = .5488

 P a g e | 390

P(x < 2 weeks): 2 weeks = 0.5 month.

P(x < 0.5 month  = 0.6 per month) =

1 - P(x > 0.5 month  = 0.6 per month)

But P(x > 0.5 month  = 0.6 per month):

Let x0 = 0.5

P(x > 0.5) = e-x = e-0.6(.5) = e-0.30 = .7408

P(x < 0.5 month) = 1 - P(x > 0.5 month) = 1 - .7408 = .2592

Average time = Expected time = µ = 1/ = 1.67 months

6.56 n = 50 p = .80

 = np = 50(.80) = 40

n  p  q  50(.80)(. 20)
 = = 2.828
 P a g e | 391

Test: µ + 3 = 40 +3(2.828) = 40 + 8.484

31.516 to 48.484 lies between 0 and 50.

It is okay to use the normal distribution to approximate this binomial

problem.

P(x < 35): correcting for continuity: x = 34.5

34.5  40
2.828
z = = -1.94

from Table A.5, the area for z = -1.94 is .4738

P(x < 35) = .5000 - .4738 = .0262

The expected value = µ = 40

P(42 < x < 47):

correction for continuity: x = 41.5 x = 47.5

41.5  40 47.5  40
2.828 2.828
z = = 0.53 z = = 2.65

from Table A.5, the area for z = 0.53 is .2019

the area for z = 2.65 is .4960

 P a g e | 392

P(42 < x < 47) = .4960 - .2019 = .2941

 P a g e | 393

6.57 µ = 2087  = 175

If 20% are less, then 30% lie between x and µ.

z.30 = -.84

x

z =

x  2087
175
-.84 =

x = 1940

If 65% are more, then 15% lie between x and µ

z.15 = -0.39

x

z =

x  2087
175
-.39 =

x = 2018.75
 P a g e | 394

If x is more than 85%, then 35% lie between x and µ.

z.35 = 1.04

x

z =

x  2087
175
1.04 =

x = 2269

6.58  = 0.8 person minute

P(x > 1 minute  = 0.8 minute):

Let x0 = 1

P(x > 1) = e-x = e-.8(1) = e-.8 = .4493

P(x > 2.5 Minutes  = 0.8 per minute):

 P a g e | 395

Let x0 = 2.5

P(x > 2.5) = e-x = e-0.8(2.5) = e-2 = .1353

6.59 µ = 2,106,774  = 50,940

P(x > 2,200,000):

2,200,000  2,106,774
50,940
z = = 1.83

from table A.5 the area for z = 1.83 is .4664

P(x > 2,200,000) = .5000 - .4664 = .0336

P(x < 2,000,000):

2,000,000  2,106,774
50,940
z = = -2.10

from table A.5 the area for z = -2.10 is .4821

P(x < 2,000,000) = .5000 - .4821 = .0179

 P a g e | 396

6.60  = 2.2 calls 30 secs.

Expected time between calls = µ = 1/ = 1/(2.2) = .4545(30 sec.) =

13.64 sec.

P(x > 1 min.  = 2.2 calls per 30 secs.):

Since Lambda and x are for different intervals,

Change Lambda to:  = 4.4 calls/1 min.

P(x > 1 min  = 4.4 calls/1 min.):

For x0 = 1: e-x = e-4.4(1) = .0123

P(x > 2 min.  = 4.4 calls/1 min.):

For x0 = 2: e-x = e-4.4(2) = e-8.8 = .0002

6.61 This is a uniform distribution with a = 11 and b = 32.

 P a g e | 397

The mean is (11 + 32)/2 = 21.5 and the standard deviation is

12
(32 - 11)/ = 6.06. Almost 81% of the time there are less than or
equal to 28

sales associates working. One hundred percent of the time there are
less than or

equal to 34 sales associates working and never more than 34. About
23.8% of

the time there are 16 or fewer sales associates working. There are 21
or fewer

sales associates working about 48% of the time.

6.62 The weight of the rods is normally distributed with a mean of 227 mg
and a

standard deviation of 2.3 mg. The probability that a rod weighs less
than or

equal to 220 mg is .0012, less than or equal to 225 mg is .1923, less

than

or equal to 227 is .5000 (since 227 is the mean), less than 231 mg is .
9590, and
 P a g e | 398

less than or equal to 238 mg is 1.000.

6.63 The lengths of cell phone calls are normally distributed with a mean
of 2.35

minutes and a standard deviation of .11 minutes. Almost 99% of the

calls are

less than or equal to 2.60 minutes, almost 82% are less than or equal
to 2.45

minutes, over 32% are less than 2.3 minutes, and almost none are less
than

2 minutes.

6.64 The exponential distribution has  = 4.51 per 10 minutes and µ =

1/4.51 =

.22173 of 10 minutes or 2.2173 minutes. The probability that there is

less than

.1 or 1 minute between arrivals is .3630. The probability that there is

less than

.2 or 2 minutes between arrivals is .5942. The probability that there

is .5 or 5
 P a g e | 399

minutes or more between arrivals is .1049. The probability that there

is more

than 1 or 10 minutes between arrivals is .0110. It is almost

certain that there

will be less than 2.4 or 24 minutes between arrivals.

Chapter 7
Sampling and Sampling Distributions

LEARNING OBJECTIVES

The two main objectives for Chapter 7 are to give you an appreciation for the
proper application of sampling techniques and an understanding of the sampling
distributions of two statistics, thereby enabling you to:

1. Determine when to use sampling instead of a census.

2. Distinguish between random and nonrandom sampling.

3. Decide when and how to use various sampling techniques.

4. Be aware of the different types of errors that can occur in a study.

5. Understand the impact of the central limit theorem on statistical

analysis.
 P a g e | 400

x p̂
6. Use the sampling distributions of and .

CHAPTER TEACHING STRATEGY

Virtually every analysis discussed in this text deals with sample data.
It is important, therefore, that students are exposed to the ways and means
that samples are gathered. The first portion of chapter 7 deals with
sampling. Reasons for sampling versus taking a census are given. Most of
these reasons are tied to the fact that taking a census costs more than
sampling if the same measurements are being gathered. Students are then
exposed to the idea of random versus nonrandom sampling. Random
sampling appeals to their concepts of fairness and equal opportunity. This
text emphasizes that nonrandom samples are non probability samples and
cannot be used in inferential analysis because levels of confidence and/or
probability cannot be assigned. It should be emphasized throughout the
discussion of sampling techniques that as future business managers (most
students will end up as some sort of supervisor/manager) students should be
aware of where and how data are gathered for studies. This will help to
assure that they will not make poor decisions based on inaccurate and poorly
gathered data.

The central limit theorem opens up opportunities to analyze data with

a host of techniques using the normal curve. Section 7.2 begins by showing
one population (randomly generated and presented in histogram form) that is
uniformly distributed and one that is exponentially distributed. Histograms of
the means for random samples of varying sizes are presented. Note that the
distributions of means “pile up” in the middle and begin to approximate the
normal curve shape as sample size increases. Note also by observing the
values on the bottom axis that the dispersion of means gets smaller and
smaller as sample size increases thus underscoring the formula for the
 P a g e | 401


n
standard error of the mean ( ). As the student sees the central limit
theorem unfold, he/she begins to see that if the sample size is large enough,
sample means can be analyzed using the normal curve regardless of the
shape of the population.

Chapter 7 presents formulas derived from the central limit theorem for
both sample means and sample proportions. Taking the time to introduce
these techniques in this chapter can expedite the presentation of material in
chapters 8 and 9.

CHAPTER OUTLINE

7.1 Sampling

Reasons for Sampling

Reasons for Taking a Census

Frame

Random Versus Nonrandom Sampling

Random Sampling Techniques

Simple Random Sampling

Stratified Random Sampling

Systematic Sampling

Cluster or Area Sampling

Nonrandom Sampling

Convenience Sampling

Judgment Sampling

Quota Sampling

Snowball Sampling
 P a g e | 402

Sampling Error

Nonsampling Errors

x
7.2 Sampling Distribution of

Sampling from a Finite Population

p̂
7.3 Sampling Distribution of

KEY TERMS

Central Limit Theorem Quota Sampling

Cluster (or Area) Sampling Random Sampling

Convenience Sampling Sample Proportion

Disproportionate Stratified Random Sampling Sampling Error

Finite Correction Factor Simple Random Sampling

Frame Snowball Sampling

Judgment Sampling Standard Error of the Mean

Nonrandom Sampling Standard Error of the Proportion

Nonrandom Sampling Techniques Stratified Random Sampling

Nonsampling Errors Systematic Sampling

Proportionate Stratified Random Sampling Two-Stage Sampling

 P a g e | 403

SOLUTIONS TO PROBLEMS IN CHAPTER 7

7.1 a) i. A union membership list for the company.

ii. A list of all employees of the company.

b) i. White pages of the telephone directory for Utica, New York.

ii. Utility company list of all customers.

c) i. Airline company list of phone and mail purchasers of tickets from

the airline during the past six months.

ii. A list of frequent flyer club members for the airline.

d) i. List of boat manufacturer's employees.

ii. List of members of a boat owners association.

e) i. Cable company telephone directory.

ii. Membership list of cable management association.

 P a g e | 404

7.4 a) Size of motel (rooms), age of motel, geographic location.

b) Gender, age, education, social class, ethnicity.

c) Size of operation (number of bottled drinks per month), number of

employees, number of different types of drinks bottled at that location,
geographic location.

d) Size of operation (sq.ft.), geographic location, age of facility, type of

process used.

7.5 a) Under 21 years of age, 21 to 39 years of age, 40 to 55 years of age,

over 55 years of age.

b) Under $1,000,000 sales per year, $1,000,000 to $4,999,999 sales per

year, $5,000,000 to $19,999,999 sales per year, $20,000,000 to
$49,000,000 per year, $50,000,000 to $99,999,999 per year, over
$100,000,000 per year.

c) Less than 2,000 sq. ft., 2,000 to 4,999 sq. ft.,

5,000 to 9,999 sq. ft., over 10,000 sq. ft.

d) East, southeast, midwest, south, southwest, west, northwest.

e) Government worker, teacher, lawyer, physician, engineer, business

person, police officer, fire fighter, computer worker.
 P a g e | 405

f) Manufacturing, finance, communications, health care, retailing, chemical,

transportation.

7.6 n = N/k = 100,000/200 = 500

7.7 N = nk = 825

7.8 k = N/n = 3,500/175 = 20

Start between 0 and 20. The human resource department probably has a list
of

company employees which can be used for the frame. Also, there might be a

company phone directory available.

 P a g e | 406

7.9 a) i. Counties

ii. Metropolitan areas

b) i. States (beside which the oil wells lie)

ii. Companies that own the wells

c) i. States

ii. Counties

7.10 Go to the district attorney's office and observe the apparent activity of
various

attorney's at work. Select some who are very busy and some who seem to
be

less active. Select some men and some women. Select some who appear
to

be older and some who are younger. Select attorneys with different ethnic

backgrounds.

7.11 Go to a conference where some of the Fortune 500 executives attend.

Approach those executives who appear to be friendly and approachable.

 P a g e | 407

7.12 Suppose 40% of the sample is to be people who presently own a personal
computer and 60% , people who do not. Go to a computer show at the city's
conference center and start interviewing people. Suppose you get enough
people who own personal computers but not enough interviews with those
who do not. Go to a mall and start interviewing people. Screen out personal
computer owners. Interview non personal computer owners until you meet
the 60% quota.

7.13 µ = 50,  = 10, n = 64

x
a) P( > 52):

x   52  50

 10
n 64
z = = 1.6

from Table A.5, Prob. = .4452

x
P( > 52) = .5000 - .4452 = .0548

x
b) P( < 51):
 P a g e | 408

x   51  50

 10
n 64
z = = 0.80

from Table A.5 prob. = .2881

x
P( < 51) = .5000 + .2881 = .7881

x
c) P( < 47):

x   47  50

 10
n 64
z = = -2.40

from Table A.5 prob. = .4918

x
P( < 47) = .5000 - .4918 = .0082

x
d) P(48.5 < < 52.4):
 P a g e | 409

x   48.5  50

 10
n 64
z = = -1.20

from Table A.5 prob. = .3849

x   52.4  50

 10
n 64
z = = 1.92

from Table A.5 prob. = .4726

x
P(48.5 < < 52.4) = .3849 + .4726 = .8575
 P a g e | 410

x
e) P(50.6 < < 51.3):

x   50.6  50

 10
n 64
z = = 0.48

from Table A.5, prob. = .1844

x   51.3  50
  1.04
 10
n 64
z =

from Table A.5, prob. = .3508

x
P(50.6 < < 51.3) = .3508 - .1844 = .1664

7.14 µ = 23.45  = 3.8

x
a) n = 10, P( > 22):
 P a g e | 411

x   22  23.45

 3.8
n 10
z = = -1.21

from Table A.5, prob. = .3869

x
P( > 22) = .3869 + .5000 = .8869

x
b) n = 4, P( > 26):

x   26  23.45

 3.8
n 4
z = = 1.34

from Table A.5, prob. = .4099

x
P( > 26) = .5000 - .4099 = .0901

x
7.15 n = 36 µ = 278 P( < 280) = .86

x
.3600 of the area lies between = 280 and µ = 278. This probability is

associated with z = 1.08 from Table A.5. Solving for  :

 P a g e | 412

x

n
z =

280  278

36
1.08 =


6
1.08 = 2

12
1.08
= = 11.11

x
7.16 n = 81  = 12 P( > 300) = .18

.5000 - .1800 = .3200 and from Table A.5, z .3200 = 0.92

Solving for µ:
 P a g e | 413

x

n
z =

300  
12
81
0.92 =

12
9
0.92 = 300 - 

1.2267 = 300 - 

µ = 300 - 1.2267 = 298.77

7.17 a) N = 1,000 n = 60 µ = 75 =6

x
P( < 76.5):

x 76.5  75

 N n 6 1000  60
n N 1 60 1000  1
z = = 2.00
 P a g e | 414

from Table A.5, prob. = .4772

x
P( < 76.5) = .4772 + .5000 = .9772

b) N = 90 n = 36 µ = 108  = 3.46

x
P(107 < < 107.7):

x 107  108


 N n 3.46 90  36
n N 1 36 90  1
z = = -2.23

from Table A.5, prob. = .4871

x 107.7  108


 N n 3.46 90  36
n N 1 36 90  1
z = = -0.67

from Table A.5, prob. = .2486

x
P(107 < < 107.7) = .4871 - .2486 = .2385

c) N = 250 n = 100 µ = 35.6  = 4.89

 P a g e | 415

x
P( > 36):

x 36  35.6

 N n 4.89 250  100
n N 1 100 250  1
z = = 1.05

from Table A.5, prob. = .3531

x
P( > 36) = .5000 - .3531 = .1469

d) N = 5000 n = 60 µ = 125  = 13.4

x
P( < 123):

x 123  125


 N n 13.4 5000  60
n N 1 60 5000  1
z = = -1.16

from Table A.5, prob. = .3770

x
P( < 123) = .5000 - .3770 = .1230
 P a g e | 416

7.18 µ = 99.9  = 30 n = 38

x
a) P( < 90):

x   90  99.9

 30
n 38
z = = -2. 03

from table A.5, area = .4788

x
P( < 90) = .5000 - .4788 = .0212

x
b) P(98 < < 105):

x   105  99.9

 30
n 38
z = = 1.05

from table A.5, area = .3531

x   98  99.9

 30
n 38
z = = -0.39
 P a g e | 417

from table A.5, area = .1517

x
P(98 < < 105) = .3531 + .1517 = .5048

x
c) P( < 112):

x   112  99.9

 30
n 38
z = = 2.49

from table A.5, area = .4936

x
P( < 112) = .5000 + .4936 = .9936

x
d) P(93 < < 96):

x   93  99.9

 30
n 38
z = = -1.42

from table A.5, area = .4222

 P a g e | 418

x   96  99.9

 30
n 38
z = = -0.80

from table A.5, area = .2881

x
P(93 < < 96) = .4222 - .2881 = .1341

7.19 N = 1500 n = 100 µ = 177,000  = 8,500

X
P( > $185,000):

X  185,000  177,000

 N n 8,500 1500  100
n N 1 100 1500  1
z = = 9.74

from Table A.5, prob. = .5000

X
P( > $185,000) = .5000 - .5000 = .0000
 P a g e | 419

x x0
7.20 µ = $65.12  = $21.45 n = 45 P( > ) = .2300

x x0
Prob. lies between and µ = .5000 - .2300 = .2700

from Table A.5, z.2700 = 0.74

x0
Solving for :

x0  

n
z =

x 0  65.12
21.45
45
0.74 =

x0 x0
2.366 = - 65.12 and = 65.12 + 2.366 = 67.486

7.21 µ = 50.4  = 11.8 n = 42

x
a) P( > 52):
 P a g e | 420

x   52  50.4

 11.8
n 42
z = = 0.88

from Table A.5, the area for z = 0.88 is .3106

x
P( > 52) = .5000 - .3106 = .1894

x
b) P( < 47.5):

x   47.5  50.4

 11.8
n 42
z = = -1.59

from Table A.5, the area for z = -1.59 is .4441

x
P( < 47.5) = .5000 - .4441 = .0559

x
c) P( < 40):

x   40  50.4

 11.8
n 42
z = = -5.71
 P a g e | 421

from Table A.5, the area for z = -5.71 is .5000

x
P( < 40) = .5000 - .5000 = .0000

d) 71% of the values are greater than 49. Therefore, 21% are between the

sample mean of 49 and the population mean, µ = 50.4.

The z value associated with the 21% of the area is -0.55

z.21 = -0.55

x

n
z =

49  50.4

42
-0.55 =

 = 16.4964
 P a g e | 422

7.22 p = .25

p̂
a) n = 110 P( < .21):

pˆ  p .21  .25

pq (.25)(. 75)
n 110
z = = -0.97

from Table A.5, prob. = .3340

p̂
P( < .21) = .5000 - .3340 = .1660

p̂
b) n = 33 P( > .24):

pˆ  p .24  .25

pq (.25)(. 75)
n 33
z = = -0.13

from Table A.5, prob. = .0517

p̂
P( > .24) = .5000 + .0517 = .5517

p̂
c) n = 59 P(.24 < < .27):
 P a g e | 423

pˆ  p .24  .25

pq (.25)(. 75)
n 59
z = = -0.18

from Table A.5, prob. = .0714

pˆ  P .27  .25

pq (.25)(. 75)
n 59
z = = 0.35

from Table A.5, prob. = .1368

p̂
P(.24 < < .27) = .0714 + .1368 = .2082
 P a g e | 424

p̂
d) n = 80 P( > .30):

pˆ  p .30  .25

pq (.25)(. 75)
n 80
z = = 1.03

from Table A.5, prob. = .3485

p̂
P( > .30) = .5000 - .3485 = .1515

p̂
e) n = 800 P( > .30):

pˆ  p .30  .25

pq (.25)(. 75)
n 800
z = = 3.27

from Table A.5, prob. = .4995

p̂
P( > .30) = .5000 - .4995 = .0005

7.23 p = .58 n = 660

 P a g e | 425

p̂
a) P( > .60):

pˆ  p .60  .58

pq (.58)(. 42)
n 660
z = = 1.04

from table A.5, area = .3508

p̂
P( > .60) = .5000 - .3508 = .1492

p̂
b) P(.55 < < .65):

pˆ  p .65  .58

pq (.58)(. 42)
n 660
z = = 3.64

from table A.5, area = .4998

pˆ  p .55  .58

pq (.58)(. 42)
n 660
z = = 1.56

from table A.5, area = .4406

 P a g e | 426

p̂
P(.55 < < .65) = .4998 + .4406 = .9404

p̂
c) P( > .57):

pˆ  p .57  .58

pq (.58)(. 42)
n 660
z = = -0.52

from table A.5, area = .1985

p̂
P( > .57) = .1985 + .5000 = .6985

p̂
d) P(.53 < < .56):

pˆ  p .56  .58 pˆ  p .53  .58

 
pq (.58)(. 42) pq (.58)(. 42)
n 660 n 660
z = = 1.04 z = =
2.60

from table A.5, area for z = 1.04 is .3508

from table A.5, area for z = 2.60 is .4953

p̂
P(.53 < < .56) = .4953 - .3508 = .1445
 P a g e | 427

p̂
e) P( < .48):

pˆ  p .48  .58

pq (.58)(. 42)
n 660
z = = -5.21

from table A.5, area = .5000

p̂
P( < .48) = .5000 - .5000 = .0000
 P a g e | 428

p̂
7.24 p = .40 P( > .35) = .8000

p̂
P(.35 < < .40) = .8000 - .5000 = .3000

from Table A.5, z.3000 = -0.84

Solving for n:

pˆ  p
pq
n
z =

.35  .40  .05

(.40)(. 60) .24
n n
-0.84 = =

 0.84 .24
 n
 .05

n
8.23 =

n = 67.73  68
 P a g e | 429

p̂ p̂0
7.25 p = .28 n = 140 P( < ) = .3000

p̂ p̂0
P( < < .28) = .5000 - .3000 = .2000

from Table A.5, z.2000 = -0.52

p̂0
Solving for :

pˆ 0  p
pq
n
z =

pˆ 0  .28
(.28)(. 72)
140
-0.52 =

p̂0
-.02 = - .28

p̂0
= .28 - .02 = .26

7.26 P(x > 150): n = 600 p = .21 x = 150

150
p̂ 600
= = .25
 P a g e | 430

pˆ  p .25  .21

pq (.21)(. 79)
n 600
z = = 2.41

from table A.5, area = .4920

P(x > 150) = .5000 - .4920 = .0080

7.27 p = .48 n = 200

a) P(x < 90):

90
p̂ 200
= = .45

pˆ  p .45  .48

pq (.48)(. 52)
n 200
z = = -0.85

from Table A.5, the area for z = -0.85 is .3023

P(x < 90) = .5000 - .3023 = .1977

b) P(x > 100):

 P a g e | 431

100
p̂ 200
= = .50

pˆ  p .50  .48

pq (.48)(. 52)
n 200
z = = 0.57

from Table A.5, the area for z = 0.57 is .2157

P(x > 100) = .5000 - .2157 = .2843

 P a g e | 432

c) P(x > 80):

80
p̂ 200
= = .40

pˆ  p .40  .48

pq (.48)(. 52)
n 200
z = = -2.26

from Table A.5, the area for z = -2.26 is .4881

P(x > 80) = .5000 + .4881 = .9881

7.28 p = .19 n = 950

p̂
a) P( > .25):

pˆ  p .25  .19

pq (.19)(.89)
n 950
z = = 4.71

from Table A.5, area = .5000

p̂
P( > .25) = .5000 - .5000 = .0000
 P a g e | 433

p̂
b) P(.15 < < .20):

pˆ  p .15  .19

pq (.19)(.81)
n 950
z = = -3.14

pˆ  p .20  .19

pq (.19)(.89)
n 950
z = = 0.79

from Table A.5, area for z = -3.14 is .4992

from Table A.5, area for z = 0.79 is .2852

p̂
P(.15 < < .20) = .4992 + .2852 = .7844

c) P(133 < x < 171):

133 171
p̂1 950 p̂2 950
= = .14 = = .18

p̂
P(.14 < < .18):
 P a g e | 434

pˆ  p .14  .19 pˆ  p .18  .19

 
pq (.19)(. 81) pq (.19)(.81)
n 950 n 950
z = = -3.93 z = =
-0.79

from Table A.5, the area for z = -3.93 is .49997

the area for z = -0.79 is .2852

P(133 < x < 171) = .49997 - .2852 = .21477

7.29 µ = 76,  = 14

x
a) n = 35, P( > 79):

x   79  76

 14
n 35
z = = 1.27

from table A.5, area = .3980

x
P( > 79) = .5000 - .3980 = .1020
 P a g e | 435

x
b) n = 140, P(74 < < 77):

x   74  76 x   77  76
 
 14  14
n 140 n 140
z = = -1.69 z = = 0.85

from table A.5, area for z = -1.69 is .4545

from table A.5, area for 0.85 is .3023

x
P(74 < < 77) = .4545 + .3023 = .7568
 P a g e | 436

x
c) n = 219, P( < 76.5):

x   76.5  76

 14
n 219
z = = 0.53

from table A.5, area = .2019

x
P( < 76.5) = .5000 + .2019 = .7019

7.30 p = .46

a) n = 60

p̂
P(.41 < < .53):

pˆ  p .53  .46

pq (.46)(. 54)
n 60
z = = 1.09

from table A.5, area = .3621

 P a g e | 437

pˆ  p .41  .46

pq (.46)(.54)
n 60
z = = -0.78

from table A.5, area = .2823

p̂
P(.41 < < .53) = .3621 + .2823 = .6444

p̂
b) n = 458 P( < .40):

p  p .4 0  .4 6

p q ( .4 6 ) ( .5 4 )
n 458
z = = -2.58

from table A.5, area = .4951

p
P( < .40) = .5000 - .4951 = .0049
 P a g e | 438

p̂
c) n = 1350 P( > .49):

pˆ  p .49  .46

pq (.46)(. 54)
n 1350
z = = 2.21

from table A.5, area = .4864

p̂
P( > .49) = .5000 - .4864 = .0136

7.31 Under 18 250(.22) = 55

18 – 25 250(.18) = 45

26 – 50 250(.36) = 90

51 – 65 250(.10) = 25

over 65 250(.14) = 35

n = 250

7.32 p = .55 n = 600 x = 298

x 298

p̂ n 600
= = .497
 P a g e | 439

p̂
P( < .497):

pˆ  p .497  .55

pq (.55)(. 45)
n 600
z = = -2.61

from Table A.5, Prob. = .4955

p̂
P( < .497) = .5000 - .4955 = .0045

No, the probability of obtaining these sample results by chance from a

population that supports the candidate with 55% of the vote is extremely low
(.0045). This is such an unlikely chance sample result that it would cause the
researcher to probably reject her claim of 55% of the vote.
 P a g e | 440

7.33 a) Roster of production employees secured from the human

resources department of the company.

b) Alpha/Beta store records kept at the headquarters of

their California division or merged files of store

records from regional offices across the state.

c) Membership list of Maine lobster catchers association.

7.34 µ = $ 17,755  = $ 650 n = 30 N = 120

x
P( < 17,500):

17,500  17,755
650 120  30
30 120  1
z = = -2.47

from Table A.5, the area for z = -2.47 is .4932

x
P( < 17,500) = .5000 - .4932 = .0068
 P a g e | 441

7.35 Number the employees from 0001 to 1250. Randomly sample from the
random number table until 60 different usable numbers are obtained. You
cannot use numbers from 1251 to 9999.

x
7.36 µ = $125 n = 32 = $110 2 = $525

x
P( > $110):

x   110  125

 525
n 32
z = = -3.70

from Table A.5, Prob.= .5000

x
P( > $110) = .5000 + .5000 = 1.0000
 P a g e | 442

x
P( > $135):

x   135  125

 525
n 32
z = = 2.47

from Table A.5, Prob.= .4932

x
P( > $135) = .5000 - .4932 = .0068

x
P($120 < < $130):

x   120  125

 525
n 32
z = = -1.23

x   130  125

 525
n 32
z = = 1.23

from Table A.5, Prob.= .3907

 P a g e | 443

x
P($120 < < $130) = .3907 + .3907 = .7814

7.37 n = 1100

a) x > 810, p = .73

x 810

p̂ n 1100
=

pˆ  p .7364  .73

pq (.73)(. 27)
n 1100
z = = 0.48

from table A.5, area = .1844

P(x > 810) = .5000 - .1844 = .3156

b) x < 1030, p = .96,

x 1030

p̂ n 1100
= = .9364
 P a g e | 444

pˆ  p .9364  .96

pq (.96)(.04)
n 1100
z = = -3.99

from table A.5, area = .49997

P(x < 1030) = .5000 - .49997 = .00003

c) p = .85

p̂
P(.82 < < .84):

pˆ  p .82  .85

pq (.85)(. 15)
n 1100
z = = -2.79

from table A.5, area = .4974

pˆ  p .84  .85

pq (.85)(. 15)
n 1100
z = = -0.93

from table A.5, area = .3238

p̂
P(.82 < < .84) = .4974 - .3238 = .1736
 P a g e | 445

7.38 1) The managers from some of the companies you are interested in

studying do not belong to the American Managers Association.

2) The membership list of the American Managers Association is not up-to-

date.

3) You are not interested in studying managers from some of the

companies belonging to the American Management Association.

4) The wrong questions are asked.

5) The manager incorrectly interprets a question.

6) The assistant accidentally marks the wrong answer.

7) The wrong statistical test is used to analyze the data.

8) An error is made in statistical calculations.

9) The statistical results are misinterpreted.

7.39 Divide the factories into geographic regions and select a few factories to
represent those regional areas of the country. Take a random sample of
employees from each selected factory. Do the same for distribution centers
and retail outlets. Divide the United States into regions of areas. Select a
few areas. Take a random sample from each of the selected area distribution
centers and retail outlets.

7.40 N = 12,080 n = 300

k = N/n = 12,080/300 = 40.27

Select every 40th outlet to assure n > 300 outlets.

 P a g e | 446

Use a table of random numbers to select a value between 0 and 40 as a

starting point.

7.41 p = .54 n = 565

a) P(x > 339):

x 339

p̂ n 565
= = .60

pˆ  p .60  .54

pq (.54)(.46)
n 565
z = = 2.86

from Table A.5, the area for z = 2.86 is .4979

P(x > 339) = .5000 - .4979 = .0021

b) P(x > 288):

x 288

p̂ n 565
= = .5097
 P a g e | 447

pˆ  p .5097  .54

pq (.54)(.46)
n 565
z = = -1.45

from Table A.5, the area for z = -1.45 is .4265

P(x > 288) = .5000 + .4265 = .9265

p̂
c) P( < .50):

pˆ  p .50  .54

pq (.54)(.46)
n 565
z = = -1.91

from Table A.5, the area for z = -1.91 is .4719

p̂
P( < .50) = .5000 - .4719 = .0281

7.42 µ = $550 n = 50  = $100

x
P( < $530):
 P a g e | 448

x   530  550

 100
n 50
z = = -1.41

from Table A.5, Prob.=.4207

P(x < $530) = .5000 - .4207 = .0793

7.43 µ = 56.8 n = 51  = 12.3

x
a) P( > 60):

x   60  56.8

 12.3
n 51
z = = 1.86

from Table A.5, Prob. = .4686

x
P( > 60) = .5000 - .4686 = .0314
 P a g e | 449

x
b) P( > 58):

x   58  56.8

 12.3
n 51
z = = 0.70

from Table A.5, Prob.= .2580

x
P( > 58) = .5000 - .2580 = .2420

x
c) P(56 < < 57):

x   56  56.8 x   57  56.8
 
 12.3  12.3
n 51 n 51
z = = -0.46 z = = 0.12

from Table A.5, Prob. for z = -0.46 is .1772

from Table A.5, Prob. for z = 0.12 is .0478

x
P(56 < < 57) = .1772 + .0478 = .2250
 P a g e | 450

x
d) P( < 55):

x   55  56.8

 12.3
n 51
z = = -1.05

from Table A.5, Prob.= .3531

x
P( < 55) = .5000 - .3531 = .1469

x
e) P( < 50):

x   50  56.8

 12.3
n 51
z = = -3.95

from Table A.5, Prob.= .5000

x
P( < 50) = .5000 - .5000 = .0000

7.45 p = .73 n = 300

a) P(210 < x < 234):

 P a g e | 451

x 210 x 234
 
p̂1 n 300 p̂2 n 300
= = .70 = = .78

pˆ  p .70  .73

pq (.73)(. 27)
n 300
z = = -1.17

pˆ  p .78  .73

pq (.73)(. 27)
n 300
z = = 1.95

from Table A.5, the area for z = -1.17 is .3790

the area for z = 1.95 is .4744

P(210 < x < 234) = .3790 + .4744 = .8534

p̂
b) P( > .78):

pˆ  p .78  .73

pq (.73)(. 27)
n 300
z = = 1.95

from Table A.5, the area for z = 1.95 is .4744

p̂
P( > .78) = .5000 - .4744 = .0256
 P a g e | 452

p̂
c) p = .73 n = 800 P( > .78):

pˆ  p .78  .73

pq (.73)(. 27)
n 800
z = = 3.19

from Table A.5, the area for z = 3.19 is .4993

p̂
P( > .78) = .5000 - .4993 = .0007
 P a g e | 453

7.46 n = 140 P(x > 35):

35
p̂ 140
= = .25 p = .22

pˆ  p .25  .22

pq (.22)(.78)
n 140
z = = 0.86

from Table A.5, the area for z = 0.86 is .3051

P(x > 35) = .5000 - .3051 = .1949

P(x < 21):

21
p̂ 140
= = .15

pˆ  p .15  .22

pq (.22)(.78)
n 140
z = = -2.00

from Table A.5, the area for z = 2.00 is .4772

P(x < 21) = .5000 - .4772 = .0228

 P a g e | 454

n = 300 p = .20

p̂
P(.18 < < .25):

pˆ  p .18  .20

pq (.20)(.80)
n 300
z = = -0.87

from Table A.5, the area for z = -0.87 is .3078

pˆ  p .25  .20

pq (.20)(.80)
n 300
z = = 2.17

from Table A.5, the area for z = 2.17 is .4850

p̂
P(.18 < < .25) = .3078 + .4850 = .7928

7.47 By taking a sample, there is potential for obtaining more detailed information.

More time can be spent with each employee. Probing questions can

be asked. There is more time for trust to be built between employee and

interviewer resulting in the potential for more honest, open answers.

With a census, data is usually more general and easier to analyze because it
is in a more standard format. Decision-makers are sometimes more
comfortable with a census because everyone is included and there is no
 P a g e | 455

sampling error. A census appears to be a better political device because the

CEO can claim that everyone in the company has had input.

7.48 p = .75 n = 150 x = 120

p̂
P( > .80):

pˆ  p .80  .75

pq (.75)(. 25)
n 150
z = = 1.41

from Table A.5, the area for z = 1.41 is .4207

p̂
P( > .80) = .5000 - .4207 = .0793

7.49 Switzerland: n = 40 µ = $ 21.24 =$3

x
P(21 < < 22):

x   21  21.24

 3
n 40
z = = -0.51
 P a g e | 456

x   22  21.24

 3
n 40
z = = 1.60

from Table A.5, the area for z = -0.51 is .1950

the area for z = 1.60 is .4452

x
P(21 < < 22) = .1950 + .4452 = .6402
 P a g e | 457

Japan: n = 35 µ = $ 22.00  = $3

x
P( > 23):

x   23  22

 3
n 35
z = = 1.97

from Table A.5, the area for z = 1.97 is .4756

x
P( > 23) = .5000 - .4756 = .0244

U.S.: n = 50 µ = $ 19.86  =$3

x
P( < 18.90):

x   18.90  19.86

 3
n 50
z = = -2.26

from Table A.5, the area for z = -2.26 is .4881

x
P( < 18.90) = .5000 - .4881 = .0119
 P a g e | 458

7.50 a) Age, Ethnicity, Religion, Geographic Region, Occupation, Urban-

Suburban-Rural, Party Affiliation, Gender

b) Age, Ethnicity, Gender, Geographic Region, Economic Class

c) Age, Ethnicity, Gender, Economic Class, Education

d) Age, Ethnicity, Gender, Economic Class, Geographic Location

7.51 µ = $281 n = 65  = $47

x
P( > $273):

x   273  281

 47
n 65
z = = -1.37

from Table A.5 the area for z = -1.37 is .4147

x
P( > $273) = .5000 + .4147 = .9147
 P a g e | 459

Chapter 8
Statistical Inference: Estimation for Single
Populations

LEARNING OBJECTIVES

The overall learning objective of Chapter 8 is to help you understand

estimating

parameters of single populations, thereby enabling you to:

1. Know the difference between point and interval estimation.

2. Estimate a population mean from a sample mean when  is known.

3. Estimate a population mean from a sample mean when  is unknown.

4. Estimate a population proportion from a sample proportion.

5. Estimate the population variance from a sample variance.

6. Estimate the minimum sample size necessary to achieve given

statistical goals.
 P a g e | 460

CHAPTER TEACHING STRATEGY

Chapter 8 is the student's introduction to interval estimation and

estimation of sample size. In this chapter, the concept of point estimate is
discussed along with the notion that as each sample changes in all likelihood
so will the point estimate. From this, the student can see that an interval
estimate may be more usable as a one-time proposition than the point
estimate. The confidence interval formulas for large sample means and
proportions can be presented as mere algebraic manipulations of formulas
developed in chapter 7 from the Central Limit Theorem.

It is very important that students begin to understand the difference

between mean and proportions. Means can be generated by averaging some
sort of measurable item such as age, sales, volume, test score, etc.
Proportions are computed by counting the number of items containing a
characteristic of interest out of the total number of items. Examples might be
proportion of people carrying a VISA card, proportion of items that are
defective, proportion of market purchasing brand A. In addition, students can
begin to see that sometimes single samples are taken and analyzed; but that
other times, two samples are taken in order to compare two brands, two
techniques, two conditions, male/female, etc.

In an effort to understand the impact of variables on confidence

intervals, it may be useful to ask the students what would happen to a
confidence interval if the sample size is varied or the confidence is increased
or decreased. Such consideration helps the student see in a different light
the items that make up a confidence interval. The student can see that
increasing the sample size reduces the width of the confidence interval, all
other things being constant, or that it increases confidence if other things are
held constant. Business students probably understand that increasing
sample size costs more and thus there are trade-offs in the research set-up.

In addition, it is probably worthwhile to have some discussion with

students regarding the meaning of confidence, say 95%. The idea is
presented in the chapter that if 100 samples are randomly taken from a
population and 95% confidence intervals are computed on each sample, that
95%(100) or 95 intervals should contain the parameter of estimation and
approximately 5 will not. In most cases, only one confidence interval is
computed, not 100, so the 95% confidence puts the odds in the researcher's
favor. It should be pointed out, however, that the confidence interval
computed may not contain the parameter of interest.
 P a g e | 461

This chapter introduces the student to the t distribution for

estimating population means when  is unknown. Emphasize that this

applies only when the population is normally distributed because it is an
assumption underlying the t test that the population is normally distributed,
albeit that this assumption is robust. The student will observe that the t
formula is essentially the same as the z formula and that it is the table that is
different. When the population is normally distributed and  is known, the z
formula can be used even for small samples.

A formula is given in chapter 8 for estimating the population variance;

and

it is here that the student is introduced to the chi-square distribution. An

assumption underlying the use of this technique is that the population is

normally

distributed. The use of the chi-square statistic to estimate the population

variance

is extremely sensitive to violations of this assumption. For this reason,

extreme

caution should be exercised in using this technique. Because of this, some

statisticians omit this technique from consideration presentation and usage.

Lastly, this chapter contains a section on the estimation of sample size.

One of the more common questions asked of statisticians is: "How large of a

sample size should I take?" In this section, it should be emphasized that

sample

size estimation gives the researcher a "ball park" figure as to how many to
sample.

The “error of estimation “ is a measure of the sampling error. It is also equal

to

the + error of the interval shown earlier in the chapter.

P a g e | 462
 P a g e | 463

CHAPTER OUTLINE

8.1 Estimating the Population Mean Using the z Statistic ( known).

Finite Correction Factor

Estimating the Population Mean Using the z Statistic when the

Sample Size is Small

Using the Computer to Construct z Confidence Intervals for the

Mean

8.2 Estimating the Population Mean Using the t Statistic ( unknown).

The t Distribution

Robustness

Characteristics of the t Distribution.

Reading the t Distribution Table

Confidence Intervals to Estimate the Population Mean Using the

t

Statistic

Using the Computer to Construct t Confidence Intervals for the

 P a g e | 464

Mean

8.3 Estimating the Population Proportion

Using the Computer to Construct Confidence Intervals of the

Population Proportion

8.4 Estimating the Population Variance

8.5 Estimating Sample Size

Sample Size When Estimating µ

Determining Sample Size When Estimating p

KEY WORDS

Bounds Point Estimate

Chi-square Distribution Robust

Degrees of Freedom(df) Sample-Size Estimation

Error of Estimation t Distribution

Interval Estimate t Value

 P a g e | 465

SOLUTIONS TO PROBLEMS IN CHAPTER 8

x
8.1 a) = 25  = 3.5 n = 60

95% Confidence z.025 = 1.96

 3.5
xz
n 60
= 25 + 1.96 = 25 + 0.89 = 24.11 < µ < 25.89

x
b) = 119.6  = 23.89 n = 75

98% Confidence z.01 = 2.33

 23.89
xz
n 75
= 119.6 + 2.33 = 119.6 ± 6.43 = 113.17 < µ <
126.03

x
c) = 3.419  = 0.974 n = 32

90% C.I. z.05 = 1.645

 P a g e | 466

 0.974
xz
n 32
= 3.419 + 1.645 = 3.419 ± .283 = 3.136 < µ <
3.702

x
d) = 56.7  = 12.1 N = 500 n = 47

80% C.I. z.10 = 1.28

 N n 12.1 500  47
xz
n N 1 47 500  1
= 56.7 + 1.28 =

56.7 ± 2.15 = 54.55 < µ < 58.85

 P a g e | 467

x
8.2 n = 36 = 211  = 23
95% C.I. z.025 = 1.96

 23
xz
n 36
= 211 ± 1.96 = 211 ± 7.51 = 203.49 < µ <
218.51

x
8.3 n = 81 = 47  = 5.89

90% C.I. z.05=1.645

 5.89
xz
n 81
= 47 ± 1.645 = 47 ± 1.08 = 45.92 < µ < 48.08

x
8.4 n = 70 2 = 49 = 90.4

x
= 90.4 Point Estimate

94% C.I. z.03 = 1.88

 49
xz
n 70
= 90.4 ± 1.88 = 90.4 ± 1.57 = 88.83 < µ < 91.97
 P a g e | 468

x
8.5 n = 39 N = 200 = 66  = 11

96% C.I. z.02 = 2.05

 N n 11 200  39
xz
n N 1 39 200  1
= 66 ± 2.05 =

66 ± 3.25 = 62.75 < µ < 69.25

x
= 66 Point Estimate

x
8.6 n = 120 = 18.72  = 0.8735

99% C.I. z.005 = 2.575

x
= 18.72 Point Estimate

 0.8735
xz
n 120
= 18.72 ± 2.575 = 8.72 ± .21 = 18.51 < µ < 18.93

x
8.7 N = 1500 n = 187 = 5.3 years  = 1.28 years

95% C.I. z.025 = 1.96

x
= 5.3 years Point Estimate
 P a g e | 469

 N n 1.28 1500  187

xz
n N 1 187 1500  1
= 5.3 ± 1.96 =

5.3 ± .17 = 5.13 < µ < 5.47

x
8.8 n = 24 = 5.625  = 3.23

90% C.I. z.05 = 1.645

 3.23
xz
n 24
= 5.625 ± 1.645 = 5.625 ± 1.085 = 4.540 < µ < 6.710

x
8.9 n = 36 = 3.306  = 1.17

98% C.I. z.01 = 2.33

 1.17
xz
n 36
= 3.306 ± 2.33 = 3.306 ± .454 = 2.852 < µ < 3.760

x
8.10 n = 36 = 2.139  = .113
 P a g e | 470

x
= 2.139 Point Estimate

90% C.I. z.05 = 1.645

 (.113)
xz
n 36
= 2.139 ± 1.645 = 2.139 ± .031 = 2.108 < µ < 2.170
 P a g e | 471

8.11 95% confidence interval n = 45

x
= 24.533 = 5.124 z = + 1.96

 5.124
xz
n 45
= 24.533 + 1.96 =

24.533 + 1.497 = 23.036 <  < 26.030

8.12 The point estimate is 0.5765. n = 41

The assumed standard deviation is 0.14

95% level of confidence: z = + 1.96

Confidence interval: 0.533647 < µ < 0.619353

Error of the estimate: 0.619353 - 0.5765 = 0.042853

x
8.13 n = 13 = 45.62 s = 5.694 df = 13 – 1 = 12

95% Confidence Interval and /2=.025

 P a g e | 472

t.025,12 = 2.179

s 5.694
xt
n 13
= 45.62 ± 2.179 = 45.62 ± 3.44 = 42.18 < µ < 49.06

x
8.14 n = 12 = 319.17 s = 9.104 df = 12 - 1 = 11

90% confidence interval

/2 = .05 t.05,11 = 1.796

s 9.104
xt
n 12
= 319.17 ± (1.796) = 319.17 ± 4.72 = 314.45 < µ <
323.89

x
8.15 n = 41 = 128.4 s = 20.6 df = 41 – 1 = 40

98% Confidence Interval

/2 = .01

t.01,40 = 2.423
 P a g e | 473

s 20.6
xt
n 41
= 128.4 ± 2.423 = 128.4 ± 7.80 = 120.6 < µ < 136.2

x
= 128.4 Point Estimate

x
8.16 n = 15 = 2.364 s2 = 0.81 df = 15 – 1 = 14

90% Confidence interval

/2 = .05

t.05,14 = 1.761

s 0.81
xt
n 15
= 2.364 ± 1.761 = 2.364 ± .409 = 1.955 < µ < 2.773

x
8.17 n = 25 = 16.088 s = .817 df = 25 – 1 = 24

99% Confidence Interval

/2 = .005

t.005,24 = 2.797
 P a g e | 474

s (.817)
xt
n 25
= 16.088 ± 2.797 = 16.088 ± .457 = 15.631 < µ < 16.545

x
= 16.088 Point Estimate
 P a g e | 475

x
8.18 n = 22 = 1,192 s = 279 df = n - 1 = 21

98% CI and /2 = .01 t.01,21 = 2.518

s 279
xt
n 22
= 1,192 + 2.518 = 1,192 + 149.78 = 1,042.22 <  <
1,341.78

The figure given by Runzheimer International falls within the confidence

interval. Therefore, there is no reason to reject the Runzheimer figure as

different from what we are getting based on this sample.

8.19 n = 20 df = 19 95% CI t.025,19 = 2.093

x
= 2.36116 s = 0.19721

0.1972
20
2.36116 + 2.093 = 2.36116 + 0.0923 = 2.26886 <  <
2.45346

Point Estimate = 2.36116

Error = 0.0923
 P a g e | 476

x
8.20 n = 28 = 5.335 s = 2.016 df = 28 – 1 = 27

90% Confidence Interval /2 = .05

t.05,27 = 1.703

s 2.016
xt
n 28
= 5.335 ± 1.703 = 5.335 + .649 = 4.686 < µ < 5.984

x
8.21 n = 10 = 49.8 s = 18.22 df = 10 – 1 = 9

95% Confidence /2 = .025 t.025,9 = 2.262

s 18.22
xt
n 10
= 49.8 ± 2.262 = 49.8 + 13.03 = 36.77 < µ < 62.83

8.22 n = 14 98% confidence /2 = .01 df = 13

t.01,13 = 2.650
 P a g e | 477

x
from data: = 152.16 s = 14.42

s 14.42
xt
n 14
confidence interval: = 152.16 + 2.65 =

152.16 + 10.21 = 141.95 <  < 162.37

The point estimate is 152.16

8.23 n = 17 df = 17 – 1 = 16 99% confidence /2 = .005

t.005,16 = 2.921

x
from data: = 8.06 s = 5.07

s 5.07
xt
n 17
confidence interval: = 8.06 + 2.921 =

8.06 + 3.59 = 4.47 <  < 11.65

 P a g e | 478

x
8.24 The point estimate is which is 25.4134 hours. The sample size is 26 skiffs.
The confidence level is 98%. The confidence interval is:
s s
x  t    x  t
n n
= 22.8124 <  < 28.0145
The error of the confidence interval is 2.6011.
 P a g e | 479

p̂
8.25 a) n = 44 =.51 99% C.I. z.005 = 2.575

pˆ  qˆ (.51)(. 49)
pˆ  z
n 44
= .51 ± 2.575 = .51 ± .194 = .316 < p< .704

p̂
b) n = 300 = .82 95% C.I. z.025 = 1.96

pˆ  qˆ (.82)(. 18)
pˆ  z
n 300
= .82 ± 1.96 = .82 ± .043 = .777 < p < .863

p̂
c) n = 1150 = .48 90% C.I. z.05 = 1.645

pˆ  qˆ (.48)(.52)
pˆ  z
n 1150
= .48 ± 1.645 = .48 ± .024 = .456 < p < .
504
 P a g e | 480

p̂
d) n = 95 = .32 88% C.I. z.06 = 1.555

pˆ  qˆ (.32)(. 68)
pˆ  z
n 95
= .32 ± 1.555 = .32 ± .074 = .246 < p < .
394

8.26 a) n = 116 x = 57 99% C.I. z.005 = 2.575

x 57

p̂ n 116
= = .49

pˆ  qˆ (.49)(. 51)
pˆ  z
n 116
= .49 ± 2.575 = .49 ± .12 = .37 < p < .61

b) n = 800 x = 479 97% C.I. z.015 = 2.17

x 479

p̂ n 800
= = .60
 P a g e | 481

pˆ  qˆ (.60)(. 40)
pˆ  z
n 800
= .60 ± 2.17 = .60 ± .038 = .562 < p < .
638

c) n = 240 x = 106 85% C.I. z.075 = 1.44

x 106

p̂ n 240
= = .44

pˆ  qˆ (.44)(. 56)
pˆ  z
n 240
= .44 ± 1.44 = .44 ± .046 = .394 < p < .
486

d) n = 60 x = 21 90% C.I. z.05 = 1.645

x 21

p̂ n 60
= = .35

pˆ  qˆ (.35)(.65)
pˆ  z
n 60
= .35 ± 1.645 = .35 ± .10 = .25 < p < .45

8.27 n = 85 x = 40 90% C.I. z.05 = 1.645

 P a g e | 482

x 40

p̂ n 85
= = .47

pˆ  qˆ (.47)(. 53)
pˆ  z
n 85
= .47 ± 1.645 = .47 ± .09 = .38 < p < .56

95% C.I. z.025 = 1.96

pˆ  qˆ (.47)(. 53)
pˆ  z
n 85
= .47 ± 1.96 = .47 ± .11 = .36 < p < .58

99% C.I. z.005 = 2.575

pˆ  qˆ (.47)(. 53)
pˆ  z
n 85
= .47 ± 2.575 = .47 ± .14 = .33 < p < .61

All other things being constant, as the confidence increased, the width of the
interval increased.

p̂
8.28 n = 1003 = .255 99% CI z.005 = 2.575
 P a g e | 483

pˆ  qˆ (.255)(.745)
pˆ  z
n 1003
= .255 + 2.575 = .255 + .035 = .220 < p < .
290

p̂
n = 10,000 = .255 99% CI z.005 = 2.575

pˆ  qˆ (.255)(.745)
pˆ  z
n 10,000
= .255 + 2.575 = .255 + .011 = .244 < p < .
266

The confidence interval constructed using n = 1003 is wider than the

confidence

interval constructed using n = 10,000. One might conclude that, all other
things

being constant, increasing the sample size reduces the width of the
confidence

interval.

p̂
8.29 n = 560 = .47 95% CI z.025 = 1.96

pˆ  qˆ (.47)(.53)
pˆ  z
n 560
= .47 + 1.96 = .47 + .0413 = .4287 < p < .5113
 P a g e | 484

p̂
n = 560 = .28 90% CI z.05 = 1.645

pˆ  qˆ (.28)(. 72)
pˆ  z
n 560
= .28 + 1.645 = .28 + .0312 = .2488 < p < .
3112

8.30 n = 1250 x = 997 98% C.I. z.01 = 2.33

x 997

p̂ n 1250
= = .80

pˆ  qˆ (.80)(.20)
pˆ  z
n 1250
= .80 ± 2.33 = .80 ± .026 = .774 < p < .826
 P a g e | 485

8.31 n = 3481 x = 927

x 927

p̂ n 3481
= = .266

p̂
a) = .266 Point Estimate

b) 99% C.I. z.005 = 2.575

pˆ  qˆ (.266)(.734)
pˆ  z
n 3481
= .266 + 2.575 = .266 ± .019 =

.247 < p < .285

8.32 n = 89 x = 48 85% C.I. z.075 = 1.44

x 48

p̂ n 89
= = .54

pˆ  qˆ (.54)(.46)
pˆ  z
n 89
= .54 ± 1.44 = .54 ± .076 = .464 < p < .616

p̂
8.33 = .63 n = 672 95% Confidence z.025 = + 1.96
 P a g e | 486

pˆ  qˆ (.63)(. 37)
pˆ  z
n 672
= .63 + 1.96 = .63 + .0365 = .5935 < p < .
6665

8.34 n = 275 x = 121 98% confidence z.01 = 2.33

x 121

p̂ n 275
= = .44

pˆ  qˆ (.44)(.56)
pˆ  z
n 275
= .44 ± 2.33 = .44 ± .07 = .37 < p < .51
 P a g e | 487

x
8.35 a) n = 12 = 28.4 s2 = 44.9 99% C.I. df =
12 – 1 = 11

2.995,11 = 2.60320 2.005,11 = 26.7569

(12  1)( 44.9) (12  1)( 44.9)

26.7569 2.60320
< 2 <

18.46 <  2 < 189.73

x
b) n=7 = 4.37 s = 1.24 s2 = 1.5376 95% C.I. df = 7 – 1 =
6

2.975,6 = 1.23734 2.025,6 = 14.4494

(7  1)(1.5376) (7  1)(1.5376)
14.4494 1.23734
< 2 <

0.64 <  2 < 7.46

x
c) n = 20 = 105 s = 32 s2 = 1024 90% C.I. df =
20 – 1 = 19

2.95,19 = 10.11701 2.05,19 = 30.1435

 P a g e | 488

(20  1)(1024) (20  1)(1024)

30.1435 10.11701
< 2 <

645.45 <  2 < 1923.10

d) n = 17 s2 = 18.56 80% C.I. df = 17 – 1 = 16

2.90,16 = 9.31224 2.10,16 = 23.5418

(17  1)(18.56) (17  1)(18.56)

23.5418 9.31224
< 2 <

12.61 <  2 < 31.89

 P a g e | 489

8.36 n = 16 s2 = 37.1833 98% C.I. df = 16-1 = 15

2.99,15 = 5.22936 2.01,15 = 30.5780

(16  1)(37.1833) (16  1)(37.1833)

30.5780 5.22936
< 2 <

18.24 <  2 < 106.66

8.37 n = 20 s = 4.3 s2 = 18.49 98% C.I. df = 20 – 1

= 19

2.99,19 = 7.63270 2.01,19 = 36.1908

(20  1)(18.49) (20  1)(18.49)

36.1908 7.63270
< 2 <

9.71 < 2 < 46.03

Point Estimate = s2 = 18.49

8.38 n = 15 s2 = 3.067 99% C.I. df = 15 – 1 = 14

2.995,14 = 4.07466 2.005,14 = 31.3194

 P a g e | 490

(15  1)(3.067) (15  1)(3.067)

31.3194 4.07466
< 2 <

1.37 < 2 < 10.54

8.39 n = 14 s2 = 26,798,241.76 95% C.I. df = 14 – 1 = 13

Point Estimate = s2 = 26,798,241.76

2.975,13 = 5.00874 2.025,13 = 24.7356

(14  1)( 26,798,241.76) (14  1)( 26,798,241.76)

24.7356 5.00874
< 2 <

14,084,038.51 <  2 < 69,553,848.45

8.40 a)  = 36 E=5 95% Confidence z.025 = 1.96

z 2 2 (1.96) 2 (36) 2

E2 52
n= = 199.15

Sample 200
 P a g e | 491

b)  = 4.13 E=1 99% Confidence z.005 = 2.575

z 2 2 ( 2.575) 2 (4.13) 2

E2 12
n= = 113.1

Sample 114

c) E = 10 Range = 500 - 80 = 420

1/4 Range = (.25)(420) = 105

90% Confidence z.05 = 1.645

z 2 2 (1.645) 2 (105) 2

E2 10 2
n = = 298.3

Sample 299

d) E=3 Range = 108 - 50 = 58

1/4 Range = (.25)(58) = 14.5

88% Confidence z.06 = 1.555

z 2 2 (1.555) 2 (14.5) 2

E2 32
n = = 56.5
 P a g e | 492

Sample 57

8.41 a) E = .02 p = .40 96% Confidence z.02 = 2.05

z 2 p  q (2.05) 2 (.40)(. 60)


E2 (.02) 2
n = = 2521.5

Sample 2522

b) E = .04 p = .50 95% Confidence z.025 = 1.96

z 2 p  q (1.96) 2 (.50)(.50)

E2 (.04) 2
n = = 600.25

Sample 601

c) E = .05 p = .55 90% Confidence z.05 = 1.645

z 2 p  q (1.645) 2 (.55)(. 45)


E2 (.05) 2
n = = 267.9

Sample 268

d) E =.01 p = .50 99% Confidence z.005 = 2.575

 P a g e | 493

z 2 p  q (2.575) 2 (.50)(.50)

E2 (.01) 2
n = = 16,576.6

Sample 16,577

8.42 E = $200  = $1,000 99% Confidence z.005 = 2.575

z 2 2 (2.575) 2 (1000) 2

E2 200 2
n = = 165.77

Sample 166
 P a g e | 494

8.43 E = $2  = $12.50 90% Confidence z.05 = 1.645

z 2 2 (1.645) 2 (12.50) 2

E2 22
n = = 105.7

Sample 106

8.44 E = $100 Range = $2,500 - $600 = $1,900

  1/4 Range = (.25)($1,900) = $475

90% Confidence z.05 = 1.645

z 2 2 (1.645) 2 (475) 2

E2 100 2
n = = 61.05

Sample 62

8.45 p = .20 q = .80 E = .02

90% Confidence, z.05 = 1.645

 P a g e | 495

z 2 p  q (1.645) 2 (.20)(. 80)


E2 (.02) 2
n = = 1082.41

Sample 1083

8.46 p = .50 q = .50 E = .05

95% Confidence, z.025 = 1.96

z 2 p  q (1.96) 2 (.50)(. 50)


E2 (.05) 2
n = = 384.16

Sample 385

8.47 E = .10 p = .50 q = .50

95% Confidence, z.025 = 1.96

z 2 p  q (1.96) 2 (.50)(. 50)


E2 (.10) 2
n = = 96.04
 P a g e | 496

Sample 97

x
8.48 = 45.6  = 7.75 n = 35

80% confidence z.10 = 1.28

 7.75
xz  45.6  1.28
n 35
= 45.6 + 1.68

43.92 <  < 47.28

94% confidence z.03 = 1.88

 7.75
xz  45.6  1.88
n 35
= 45.6 + 2.46

43.14 <  < 48.06

98% confidence z.01 = 2.33

 P a g e | 497

 7.75
xz  45.6  2.33
n 35
= 45.6 + 3.05

42.55 <  < 48.65

 P a g e | 498

x
8.49 = 12.03 (point estimate) s = .4373 n = 10 df = 9

For 90% confidence: /2 = .05 t.05,9= 1.833

s (.4373)
xt  12.03  1.833
n 10
= 12.03 + .25

11.78 <  < 12.28

For 95% confidence: /2 = .025 t.025,9 = 2.262

s (.4373)
xt  12.03  2.262
n 10
= 12.03 + .31

11.72 <  < 12.34

For 99% confidence: /2 = .005 t.005,9 = 3.25

s (.4373)
xt  12.03  3.25
n 10
= 12.03 + .45

11.58 <  < 12.48

 P a g e | 499

8.50 a) n = 715 x = 329 95% confidence z.025 = 1.96

329
pˆ 
715
= .46

pˆ  qˆ (.46)(. 54)
pˆ  z  .46  1.96
n 715
= .46 + .0365

.4235 < p < .4965

p̂
b) n = 284 = .71 90% confidence z.05 = 1.645

pˆ  qˆ (.71)(.29)
pˆ  z  .71  1.645
n 284
= .71 + .0443

.6657 < p < .7543

p̂
c) n = 1250 = .48 95% confidence z.025 = 1.96

pˆ  qˆ (.48)(. 52)
pˆ  z  .48  1.96
n 1250
= .48 + .0277

.4523 < p < .5077

 P a g e | 500

d) n = 457 x = 270 98% confidence z.01 = 2.33

270
pˆ 
457
= .591

pˆ  qˆ (.591)(.409)
pˆ  z  .591  2.33
n 457
= .591 + .0536

.5374 < p < .6446

8.51 n = 10 s = 7.40045 s2 = 54.7667 df = 10 – 1 = 9

90% confidence, /2 = .05 1 - /2 = .95

2.95,9 = 3.32512 2.05,9 = 16.9190

(10  1)(54.7667) (10  1)(54.7667)

16.9190 3.32512
< 2 <

29.133 < 2 < 148.235

 P a g e | 501

95% confidence, /2 = .025 1 - /2 = .975

2.975,9 = 2.70039 2.025,9 = 19.0228

(10  1)(54.7667) (10  1)(54.7667)

19.0228 2.70039
< 2 <

25.911 <  2 < 182.529

 P a g e | 502

8.52 a)  = 44 E=3 95% confidence z.025 = 1.96

z 2 2 (1.96) 2 (44) 2

E2 32
n = = 826.4

Sample 827

b) E=2 Range = 88 - 20 = 68

use  = 1/4(range) = (.25)(68) = 17

90% confidence z.05 = 1.645

z 2 2 (1.645) 2 (17) 2

E2 22
= 195.5

Sample 196

c) E = .04 p = .50 q = .50

98% confidence z.01 = 2.33

z 2 p  q ( 2.33) 2 (.50)(. 50)


E2 (.04) 2
= 848.3
 P a g e | 503

Sample 849

d) E = .03 p = .70 q = .30

95% confidence z.025 = 1.96

z 2 p  q (1.96) 2 (.70)(. 30)


E2 (.03) 2
= 896.4

Sample 897
 P a g e | 504

x
8.53 n = 17 = 10.765 s = 2.223 df = 17 - 1 = 16

99% confidence /2 = .005 t.005,16 = 2.921

s 2.223
xt  10.765  2.921
n 17
= 10.765 + 1.575

9.19 < µ < 12.34

8.54 p = .40 E=.03 90% Confidence z.05 = 1.645

z 2 p  q (1.645) 2 (.40)(. 60)


E2 (.03) 2
n = = 721.61

Sample 722

8.55 n = 17 s2 = 4.941 99% C.I. df = 17 – 1 = 16

2.995,16 = 5.14216 2.005,16 = 34.2671

(17  1)( 4.941) (17  1)( 4.941)

34.2671 5.14216
< 2 <
 P a g e | 505

2.307 <  2 < 15.374

x
8.56 n = 45 = 213  = 48

98% Confidence z.01 = 2.33

 48
xz  213  2.33
n 45
= 213 ± 16.67

196.33 < µ < 229.67

 P a g e | 506

x
8.57 n = 39 = 37.256  = 3.891

90% confidence z.05 = 1.645

 3.891
xz  37.256  1.645
n 39
= 37.256 ± 1.025

36.231 < µ < 38.281

8.58  = 6 E=1 98% Confidence z.98 = 2.33

z 2 2 (2.33) 2 (6) 2

E2 12
n = = 195.44

Sample 196

8.59 n = 1,255 x = 714 95% Confidence z.025 = 1.96

714
pˆ 
1255
= .569
 P a g e | 507

pˆ  qˆ (.569)(.431)
pˆ  z  .569  1.96
n 1,255
= .569 ± .027

.542 < p < .596

 P a g e | 508

x  128
8.60 n = 41 s = 21 98% C.I. df = 41 – 1 =
40

t.01,40 = 2.423

Point Estimate = $128

s 21
xt  128  2.423
n 41
= 128 + 7.947

120.053 <  < 135.947

Interval Width = 135.947 – 120.053 = 15.894

x
8.61 n = 60 = 6.717  = 3.06 N = 300

98% Confidence z.01 = 2.33

s N n 3.06 300  60
xz  6.717  2.33
n N 1 60 300  1
=
 P a g e | 509

6.717 ± 0.825

5.892 < µ < 7.542

8.62 E = $20 Range = $600 - $30 = $570

1/4 Range = (.25)($570) = $142.50

95% Confidence z.025 = 1.96

z 2 2 (1.96) 2 (142.50) 2

E2 20 2
n = = 195.02

Sample 196
 P a g e | 510

8.63 n = 245 x = 189 90% Confidence z.05= 1.645

x 189
pˆ  
n 245
= .77

pˆ  qˆ (.77)(. 23)
pˆ  z  .77  1.645
n 245
= .77 ± .044

.726 < p < .814

8.64 n = 90 x = 30 95% Confidence z.025 = 1.96

x 30
pˆ  
n 90

= .33

pˆ  qˆ (.33)(.67)
pˆ  z  .33  1.96
n 90
= .33 ± .097

.233 < p < .427

x
8.65 n = 12 = 43.7 s2 = 228 df = 12 – 1 = 11 95% C.I.
 P a g e | 511

t.025,11 = 2.201

s 228
xt  43.7  2.201
n 12
= 43.7 + 9.59

34.11 <  < 53.29

2.99,11 = 3.05350 2.01,11 = 24.7250

(12  1)( 228) (12  1)( 228)

24.7250 3.05350
< 2 <

101.44 <  2 < 821.35

 P a g e | 512

x
8.66 n = 27 = 4.82 s = 0.37 df = 26

95% CI: t.025,26 = 2.056

s 0.37
xt  4.82  2.056
n 27
= 4.82 + .1464

4.6736 < µ < 4.9664

Since 4.50 is not in the interval, we are 95% confident that µ does
not

equal 4.50.

x
8.67 n = 77 = 2.48  = 12

95% Confidence z.025 = 1.96

 12
xz  2.48  1.96
n 77
= 2.48 ± 2.68

-0.20 < µ < 5.16

The point estimate is 2.48

 P a g e | 513

The interval is inconclusive. It says that we are 95% confident that the
average arrival time is somewhere between .20 of a minute (12 seconds)
early and 5.16 minutes late. Since zero is in the interval, there is a possibility
that, on average, the flights are on time.

p̂
8.68 n = 560 =.33

99% Confidence z.005= 2.575

pˆ  qˆ (.33)(. 67)
pˆ  z  .33  2.575
n 560
= .33 ± .05

.28 < p < .38

 P a g e | 514

8.69 p = .50 E = .05 98% Confidence z.01 = 2.33

z 2 p  q (2.33) 2 (.50)(. 50)


E2 (.05) 2
= 542.89

Sample 543

x
8.70 n = 27 = 2.10 s = 0.86 df = 27 - 1 = 26

98% confidence /2 = .01 t.01,26 = 2.479

s 0.86
xt  2.10  2.479
n 27
= 2.10 ± 0.41

1.69 < µ < 2.51

8.71 n = 23 df = 23 – 1 = 22 s = .0631455 90% C.I.

2.95,22 = 12.33801 2.05,22 = 33.9245

 P a g e | 515

(23  1)(. 0631455) 2 (23  1)(.0631455) 2

33.9245 12.33801
< 2 <

.0026 <  2 < .0071

x
8.72 n = 39 = 1.294  = 0.205 99% Confidence
z.005 = 2.575

 0.205
xz  1.294  2.575
n 39
= 1.294 ± .085

1.209 < µ < 1.379

 P a g e | 516

8.73 The sample mean fill for the 58 cans is 11.9788 oz. with a standard deviation
of

.0536 oz. The 99% confidence interval for the population fill is 11.9607 oz. to

11.9969 oz. which does not include 12 oz. We are 99% confident that the
population mean is not 12 oz., indicating that the machine may be under
filling

the cans.

8.74 The point estimate for the average length of burn of the new bulb is 2198.217

hours. Eighty-four bulbs were included in this study. A 90% confidence

interval

can be constructed from the information given. The error of the confidence

interval is + 27.76691. Combining this with the point estimate yields the
90%

confidence interval of 2198.217 + 27.76691 = 2170.450 < µ < 2225.984.

8.75 The point estimate for the average age of a first time buyer is 27.63 years.
The

sample of 21 buyers produces a standard deviation of 6.54 years. We are

98%

confident that the actual population mean age of a first-time home buyer is

between 24.0222 years and 31.2378 years.

 P a g e | 517

8.76 A poll of 781 American workers was taken. Of these, 506 drive their cars to
work. Thus, the point estimate for the population proportion is 506/781 = .
647887. A 95% confidence interval to estimate the population proportion
shows that we are 95% confident that the actual value lies between .61324
and .681413. The error of this interval is + .0340865.

Chapter 9
Statistical Inference:
Hypothesis Testing for Single Populations

LEARNING OBJECTIVES

The main objective of Chapter 9 is to help you to learn how to test hypotheses on single
populations, thereby enabling you to:

1. Understand the logic of hypothesis testing and know how to establish

null and alternate hypotheses.

2. Understand Type I and Type II errors and know how to solve for Type II
errors.
3. Know how to implement the HTAB system to test hypotheses.

4. Test hypotheses about a single population mean when  is known.

5. Test hypotheses about a single population mean when  is unknown.

6. Test hypotheses about a single population proportion.

7. Test hypotheses about a single population variance.

P a g e | 518
 Page | 1

CHAPTER TEACHING STRATEGY

For some instructors, this chapter is the cornerstone of the first

statistics course.

Hypothesis testing presents the logic in which ideas, theories, etc., are
scientifically

examined. The student can be made aware that much of the development of concepts to
this point including sampling, level of data measurement, descriptive tools such as mean
and standard deviation, probability, and distributions pave the way for testing hypotheses.
Often students (and instructors) will say "Why do we need to test this hypothesis when
we can make a decision by examining the data?" Sometimes it is true that examining the
data could allow hypothesis decisions to be made. However, by using the methodology
and structure of hypothesis testing even in "obvious" situations, the researcher has added
credibility and rigor to his/her findings. Some statisticians actually report findings in a
court of law as an expert witness. Others report their findings in a journal, to the public,
to the corporate board, to a client, or to their manager. In each case, by using the
hypothesis testing method rather than a "seat of the pants" judgment, the researcher
stands on a much firmer foundation by using the principles of hypothesis testing and
random sampling. Chapter 9 brings together many of the tools developed to this point
and formalizes a procedure for testing hypotheses.

The statistical hypotheses are set up as to contain all possible

decisions. The

two-tailed test always has = and  in the null and alternative hypothesis.
One-tailed tests are presented with = in the null hypothesis and either > or <
in the alternative hypothesis. If in doubt, the researcher should use a two-
tailed test. Chapter 9 begins with a two-tailed test example. Often that
which the researcher wants to demonstrate true or prove true is set up as the
alternative hypothesis. The null hypothesis is that the new theory or idea is
not true, the status quo is still true, or that there is no difference. The null
hypothesis is assumed to be true before the process begins. Some
researchers liken this procedure to a court of law where the defendant is
presumed innocent (assume null is true - nothing has happened). Evidence is
brought before the judge or jury. If enough evidence is presented, then the
null hypothesis (defendant innocent) can no longer be accepted or assumed
true. The null hypothesis is rejected as not true and the alternate hypothesis
is accepted as true by default. Emphasize that the researcher needs to make
a decision after examining the observed statistic.
 Page | 2

Some of the key concepts in this chapter are one-tailed and two-tailed
test and Type I and Type II error. In order for a one-tailed test to be
conducted, the problem must include some suggestion of a direction to be
tested. If the student sees such words as greater, less than, more than,
higher, younger, etc., then he/she knows to use a one-tail test. If no direction
is given (test to determine if there is a "difference"), then a two-tailed test is
called for. Ultimately, students will see that the only effect of using a one-
tailed test versus a two-tailed test is on the critical table value. A one-tailed
test uses all of the value of alpha in one tail. A two-tailed test splits alpha
and uses alpha/2 in each tail thus creating a critical value that is further out
in the distribution. The result is that (all things being the same) it is more
difficult to reject the null hypothesis with a two-tailed test. Many computer
packages such as MINITAB include in the results a p-value. If you

designate that the hypothesis test is a two-tailed test, the computer will
double the p-value so that it can be compared directly to alpha.

In discussing Type I and Type II errors, there are a few things to

consider. Once a decision is made regarding the null hypothesis, there is a
possibility that the decision is correct or that an error has been made. Since
the researcher virtually never knows for certain whether the null hypothesis
was actually true or not, a probability of committing one of these errors can
be computed. Emphasize with the students that a researcher can never
commit a Type I error and a Type II error at the same time. This is so because
a Type I error can only be committed when the null hypothesis is rejected and
a Type II error can only be committed when the decision is to not reject the
null hypothesis. Type I and Type II errors are important concepts for
managerial students to understand even beyond the realm of statistical
hypothesis testing. For example, if a manager decides to fire or not fire an
employee based on some evidence collected, he/she could be committing a
Type I or a Type II error depending on the decision. If the production manager
decides to stop the production line because of evidence of faulty raw
materials, he/she might be committing a Type I error. On the other hand, if
the manager fails to shut the production line down even when faced with
evidence of faulty raw materials, he/she might be committing a Type II error.

The student can be told that there are some widely accepted values for
alpha (probability of committing a Type I error) in the research world and that
 Page | 3

a value is usually selected before the research begins. On the other hand,
since the value of Beta (probability of committing a Type II error) varies with
every possible alternate value of the parameter being tested, Beta is usually
examined and computed over a range of possible values of that parameter.
As you can see, the concepts of hypothesis testing are difficult and represent
higher levels of learning (logic, transfer, etc.). Student understanding of
these concepts will improve as you work your way through the techniques in
this chapter and in chapter 10.

CHAPTER OUTLINE

9.1 Introduction to Hypothesis Testing

Types of Hypotheses

Research Hypotheses

Statistical Hypotheses

Substantive Hypotheses

Using the HTAB System to Test Hypotheses

Rejection and Non-rejection Regions

Type I and Type II errors

9.2 Testing Hypotheses About a Population Mean Using the z Statistic (

known)

Testing the Mean with a Finite Population

Using the p-Value Method to Test Hypotheses

Using the Critical Value Method to Test Hypotheses

Using the Computer to Test Hypotheses about a Population Mean

Using
 Page | 4

the z Statistic

9.3 Testing Hypotheses About a Population Mean Using the t Statistic (

unknown)

Using the Computer to Test Hypotheses about a Population Mean

Using

the t Test

9.4 Testing Hypotheses About a Proportion

Using the Computer to Test Hypotheses about a Population

Proportion

9.5 Testing Hypotheses About a Variance

9.6 Solving for Type II Errors

Some Observations About Type II Errors

Operating Characteristic and Power Curves

Effect of Increasing Sample Size on the Rejection Limits

 Page | 5

KEY TERMS

Alpha( ) One-tailed Test

Alternative Hypothesis Operating-Characteristic Curve

(OC)

Beta( ) p-Value Method

Critical Value Power

Critical Value Method Power Curve

Hypothesis Rejection Region

Hypothesis Testing Research Hypothesis

Level of Significance Statistical Hypothesis

Nonrejection Region Substantive Result

Null Hypothesis Two-Tailed Test

Observed Significance Level Type I Error

Observed Value

Type II Error

SOLUTIONS TO PROBLEMS IN CHAPTER 9

9.1 a) Ho: µ = 25

Ha: µ  25
 Page | 6

x
= 28.1 n = 57  = 8.46  = .01

For two-tail, /2 = .005 zc = 2.575

x   28.1  25

 8.46
n 57
z= = 2.77

observed z = 2.77 > zc = 2.575

Reject the null hypothesis

b) from Table A.5, inside area between z = 0 and z = 2.77 is .4972

p-value = .5000 - .4972 = .0028

Since the p-value of .0028 is less than /2 = .005, the decision is to:

Reject the null hypothesis

 Page | 7

c) critical mean values:

xc  

n
zc =

x c  25
8.46
57
± 2.575 =

x
c = 25 ± 2.885

x
c = 27.885 (upper value)

x
c = 22.115 (lower value)

9.2 Ho: µ = 7.48

Ha: µ < 7.48

x
= 6.91 n = 24  = 1.21  =.01
 Page | 8

For one-tail,  = .01 zc = -2.33

x   6.91  7.48

 1.21
n 24
z = = -2.31

observed z = -2.31 > zc = -2.33

Fail to reject the null hypothesis

 Page | 9

9.3 a)Ho: µ = 1,200

Ha: µ > 1,200

x
= 1,215 n = 113  = 100  = .10

For one-tail,  = .10 zc = 1.28

x   1,215  1,200

 100
n 113
z = = 1.59

observed z = 1.59 > zc = 1.28

Reject the null hypothesis

b) Probability > observed z = 1.59 is .5000 - .4441 = .0559 (the p-

value) which is

less than  = .10.

Reject the null hypothesis.

c) Critical mean value:

xc  

n
zc =
 P a g e | 10

x c  1,200
100
113
1.28 =

x
c = 1,200 + 12.04

x x
Since the observed = 1,215 is greater than the critical = 1212.04,
the decision is to reject the null hypothesis.
 P a g e | 11

9.4 Ho: µ = 82

Ha: µ < 82

x
= 78.125 n = 32  = 9.184  = .01

z.01 = -2.33

x   78.125  82

 9.184
n 32
z= = -2.39

Since observed z = -2.39 < z.01 = -2.33

Reject the null hypothesis

Statistically, we can conclude that urban air soot is significantly lower.

From a business and community point-of-view, assuming that the
sample result is representative of how the air actually is now; is a
reduction of suspended particles from 82 to 78.125 really an important
reduction in air pollution (is it substantive)? Certainly it marks an
important first step and perhaps a significant start. Whether or not it
would really make a difference in the quality of life for people in the
city of St. Louis remains to be seen. Most likely, politicians and city
chamber of commerce folks would jump on such results as indications
of improvement in city conditions.

9.5 H0:  = $424.20

 P a g e | 12

Ha:   $424.20

x
= $432.69 n = 54  = $33.90  = .05

2-tailed test, /2 = .025 z.025 = + 1.96

x   432.69  424.20

 33.90
n 54
z = = 1.84

Since the observed z = 1.85 < z.025 = 1.96, the decision is to fail to
reject the

null hypothesis.
 P a g e | 13

9.6 H0:  = $62,600

Ha:  < $62,600

x
= $58,974 n = 18  = $7,810  = .01

1-tailed test,  = .01 z.01 = -2.33

x   58,974  62,600

 7,810
n 18
z = = -1.97

Since the observed z = -1.97 > z.01 = -2.33, the decision is to fail to
reject the

null hypothesis.

9.7 H0:  = 5

Ha:   5

x
= 5.0611 n = 42 N = 650  = 0.2803 =.
10

2-tailed test, /2 = .05 z.05 = + 1.645

 P a g e | 14

x 5.0611  5

 N n 0.2803 650  42
n N 1 42 650  1
z = = 1.46

Since the observed z = 1.46 < z.05 = 1.645, the decision is to fail to
reject the

null hypothesis.

9.8 Ho: µ = 18.2

Ha: µ < 18.2

x
= 15.6 n = 32  = 2.3  = .10

For one-tail,  = .10, z.10 = -1.28

x   15.6  18.2

 2 .3
n 32
z = = -6.39

Since the observed z = -6.39 < z.10 = -1.28, the decision is to

Reject the null hypothesis

9.9 Ho: µ = $4,292

Ha: µ < $4,292

 P a g e | 15

x
= $4,008 n = 55 = $386  = .01

For one-tailed test,  = .01, z.01 = -2.33

x   $4,008  $4,292

 $386
n 55
z = = -5.46

Since the observed z = -5.46 < z.01 = -2.33, the decision is to

Reject the null hypothesis

The CEO could use this information as a way of discrediting the

Runzheimer study and using her own figures in recruiting people and in
discussing relocation options. In such a case, this could be a
substantive finding. However, one must ask if the difference between
$4,292 and $4,008 is really an important difference in monthly rental
expense. Certainly, Paris is expensive either way. However, an almost
$300 difference in monthly rental cost is a nontrivial amount for most
people and therefore might be considered substantive.

9.10 Ho: µ = 123

Ha: µ > 123

 = .05 n = 40 40 people were sampled

 P a g e | 16

x
= 132.36 s = 27.68

This is a one-tailed test. Since the p-value = .016, we

reject the null hypothesis at  = .05.

The average water usage per person is greater than 123 gallons.
 P a g e | 17

x
9.11 n = 20 = 16.45 s = 3.59 df = 20 - 1 = 19  = .05

Ho: µ = 16

Ha: µ  16

For two-tail test, /2 = .025, critical t.025,19 = ±2.093

x   16.45  16

s 3.59
n 20
t = = 0.56

Observed t = 0.56 < t.025,19 = 2.093

The decision is to Fail to reject the null hypothesis

x
9.12 n = 51 = 58.42 s2 = 25.68 df = 51 - 1 = 50 
= .01

Ho: µ = 60

Ha: µ < 60

For one-tail test,  = .01 critical t.01,50 = -2.403

 P a g e | 18

x   58.42  60

s 25.68
n 51
t = = -2.23

Observed t = -2.23 > t.01,7 = -2.403

The decision is to Fail to reject the null hypothesis

 P a g e | 19

x
9.13 n = 11 = 1,235.36 s = 103.81 df = 11 - 1 =
10  = .05

Ho: µ = 1,160

Ha: µ > 1,160

or one-tail test,  = .05 critical t.05,10 = 1.812

x   1,236.36  1,160

s 103.81
n 11
t = = 2.44

Observed t = 2.44 > t.05,10 = 1.812

The decision is to Reject the null hypothesis

x
9.14 n = 20 = 8.37 s = .1895 df = 20-1 = 19
 = .01

Ho: µ = 8.3

Ha: µ  8.3

For two-tail test, /2 = .005 critical t.005,19 = ±2.861

 P a g e | 20

x   8.37  8.3

s .1895
n 20
t = = 1.65

Observed t = 1.65 < t.005,19 = 2.861

The decision is to Fail to reject the null hypothesis

x
9.15 n = 12 = 1.85083 s = .02353 df = 12 - 1 = 11
 = .10

H0: µ = 1.84

Ha: µ  1.84

For a two-tailed test, /2 = .05 critical t.05,11 = 1.796

x   1.85083  1.84

s .02353
n 12
t = = 1.59

Since t = 1.59 < t11,.05 = 1.796,

The decision is to fail to reject the null hypothesis.

x
9.16 n = 25 = 3.1948 s = .0889 df = 25 - 1 = 24  = .01
 P a g e | 21

Ho: µ = $3.16

Ha: µ > $3.16

For one-tail test, = .01 Critical t.01,24 = 2.492

x   3.1948  3.16

s .0889
n 25
t = = 1.96

Observed t = 1.96 < t.01,24 = 2.492

The decision is to Fail to reject the null hypothesis

x
9.17 n = 19 = $31.67 s = $1.29 df = 19 – 1 = 18
 = .05

H0:  = $32.28

Ha:   $32.28

Two-tailed test, /2 = .025 t.025,18 = + 2.101

x   31.67  32.28

s 1.29
n 19
t = = -2.06
 P a g e | 22

The observed t = -2.06 > t.025,18 = -2.101,

The decision is to fail to reject the null hypothesis

x
9.18 n = 61 = 3.72 s = 0.65 df = 61 – 1 = 60
 = .01

H0:  = 3.51

Ha:  > 3.51

One-tailed test,  = .01 t.01,60 = 2.390

x   3.72  3.51

s 0.65
n 61
t = = 2.52

The observed t = 2.52 > t.01,60 = 2.390,

The decision is to reject the null hypothesis

 P a g e | 23

x
9.19 n = 22 = 1031.32 s = 240.37 df = 22 – 1 = 21
 = .05

H0:  = 1135


Ha:  1135

Two-tailed test, /2= .025 t.025,21 = +2.080

x   1031.32  1135

s 240.37
n 22
t = = -2.02

The observed t = -2.02 > t.025,21 = -2.080,

The decision is to fail to reject the null hypothesis

x
9.20 n = 12 = 42.167 s = 9.124 df = 12 – 1 = 11
 = .01

H0:  = 46

Ha:  < 46

One-tailed test, = .01 t.01,11 = -2.718

 P a g e | 24

x   42.167  46

s 9.124
n 12
t = = -1.46

The observed t = -1.46 > t.01,11 = -2.718,

The decision is to fail to reject the null hypothesis

x
9.21 n = 26 = 19.534 minutes s = 4.100 minutes 
= .05

H0:  = 19

Ha:   19

Two-tailed test, /2 = .025, critical t value = + 2.06

Observed t value = 0.66. Since the observed t = 0.66 < critical t

value = 2.06,

The decision is to fail to reject the null hypothesis.

Since the Excel p-value = .256 > /2 = .025 and MINITAB p-value
=.513 > .05, the decision is to fail to reject the null hypothesis.

She would not conclude that her city is any different from the
ones in the

national survey.
 P a g e | 25

9.22 Ho: p = .45

Ha: p > .45

p̂
n = 310 = .465  = .05

For one-tail,  = .05 z.05 = 1.645

pˆ  p .465  .45

pq (.45)(. 55)
n 310
z = = 0.53

observed z = 0.53 < z.05 = 1.645

The decision is to Fail to reject the null hypothesis

9.23 Ho: p = 0.63

Ha: p < 0.63

x 55
pˆ  
n 100
n = 100 x = 55 = .55

For one-tail,  = .01 z.01 = -2.33

 P a g e | 26

pˆ  p .55  .63

pq (.63)(. 37)
n 100
z = = -1.66

observed z = -1.66 > zc = -2.33

The decision is to Fail to reject the null hypothesis

 P a g e | 27

9.24 Ho: p = .29

Ha: p  .29

x 207
pˆ  
n 740
n = 740 x = 207 = .28  = .05

For two-tail, /2 = .025 z.025 = ±1.96

pˆ  p .28  .29

pq (.29)(. 71)
n 740
z = = -0.60

observed z = -0.60 > zc = -1.96

The decision is to Fail to reject the null hypothesis

p-Value Method:

z = -0.60

from Table A.5, area = .2257

Area in tail = .5000 - .2257 = .2743 which is the p-value

 P a g e | 28

Since the p-value = .2743 > /2 = .025, the decision is to Fail to
reject the null

hypothesis

Solving for critical values:

pˆ c  p
pq
n
z =

pˆ c  .29
(.29)(. 71)
740
±1.96 =

p̂c
= .29 ± .033

.257 and .323 are the critical values

p̂
Since = .28 is not outside critical values in tails, the decision is to Fail
to reject

the null hypothesis

 P a g e | 29

9.25 Ho: p = .48

Ha: p  .48

n = 380 x = 164  = .01 /2 = .005 z.005 = +2.575

x 164
pˆ  
n 380
= .4316

pˆ  p .4316  .48

pq (.48)(. 52)
n 380
z = = -1.89

Since the observed z = -1.89 is greater than z.005= -2.575, The decision
is to fail to reject the null hypothesis. There is not enough
evidence to declare that the proportion is any different than .48.

9.26 Ho: p = .79

Ha: p < .79

n = 415 x = 303  = .01 z.01 = -2.33

x 303
pˆ  
n 415
= .7301
 P a g e | 30

pˆ  p 7301  .79

pq (.79)(.21)
n 415
z = = -3.00

Since the observed z = -3.00 is less than z.01= -2.33, The decision is to
reject the null hypothesis.
 P a g e | 31

9.27 Ho: p = .31

Ha: p  .31

n = 600 x = 200  = .10 /2 = .05 z.005 = +1.645

x 200
pˆ  
n 600
= .3333

pˆ  p .3333  .31

pq (.31)(. 69)
n 600
z = = 1.23

Since the observed z = 1.23 is less than z.005= 1.645, The decision is to
fail to reject the null hypothesis. There is not enough evidence to
declare that the proportion is any different than .31.

Ho: p = .24

Ha: p < .24

n = 600 x = 130  = .05 z.05 = -1.645

x 130
pˆ  
n 600
= .2167
 P a g e | 32

pˆ  p .2167  .24

pq (.24)(.76)
n 600
z = = -1.34

Since the observed z = -1.34 is greater than z.05= -1.645, The decision
is to fail to reject the null hypothesis. There is not enough
evidence to declare that the proportion is less than .24.
 P a g e | 33

9.28 Ho: p = .18

Ha: p > .18

p̂
n = 376 = .22  = .01

one-tailed test, z.01 = 2.33

pˆ  p .22  .18

pq (.18)(. 82)
n 376
z = = 2.02

Since the observed z = 2.02 is less than z.01= 2.33, The decision is to
fail to reject the null hypothesis. There is not enough evidence to
declare that the proportion is greater than .18.

9.29 Ho: p = .32

Ha: p < .32

x 22
pˆ  
n 118
n = 118 x = 22 = .1864  = .05

For one-tailed test, z.05 = -1.645

 P a g e | 34

pˆ  p .1864  .32

pq (.32)(.68)
n 118
z = = -3.11

Observed z = -3.11 < z.05 –1.645

Since the observed z = -3.11 is less than z.05= -1.645, the decision is to
reject the null hypothesis.
 P a g e | 35

9.30 Ho: p = .47

Ha: p  .47

n = 67 x = 40  = .05 /2 = .025

For a two-tailed test, z.025 = +1.96

x 40
pˆ  
n 67
= .597

pˆ  p .597  .47

pq (.47)(.53)
n 67
z = = 2.08

Since the observed z = 2.08 is greater than z.025= 1.96, The decision is
to reject the null hypothesis.

9.31 a) H0: 2 = 20  = .05 n = 15 df = 15 – 1 = 14 s2 =

32

Ha: 2 > 20

2.05,14 = 23.6848
 P a g e | 36

(15  1)(32)
20
2 = = 22.4

Since 2 = 22.4 < 2.05,14 = 23.6848, the decision is to fail to reject

the null

hypothesis.

b) H0: 2 = 8.5  = .10 /2 = .05 n = 22 df = n-1 = 21

2
s = 17

Ha: 2  8.5

2.05,21 = 32.6706

(22  1)(17)
8.5
2 = = 42

Since 2 = 42 > 2.05,21 = 32.6706, the decision is to reject the null

hypothesis.

c) H0: 2 = 45  = .01 n=8 df = n – 1 = 7 s = 4.12

Ha: 2 < 45

2 .01,7 = 18.4753
 P a g e | 37

(8  1)( 4.12) 2
45
2 = = 2.64

Since 2 = 2.64 < 2.01,7 = 18.4753, the decision is to fail to reject

the null

hypothesis.

d) H 0: 2 = 5  = .05 /2 = .025 n = 11 df = 11 – 1 = 10

s2 = 1.2

Ha: 2  5

2.025,10 = 20.4832 2.975,10 = 3.24696

(11  1)(1.2)
5
2 = = 2.4

Since 2 = 2.4 < 2.975,10 = 3.24696, the decision is to reject the null
hypothesis.

9.32 H0: 2 = 14  = .05 /2 = .025 n = 12 df = 12 – 1 = 11

s2 = 30.0833

Ha: 2  14

2.025,11 = 21.9200 2.975,11 = 3.81574

 P a g e | 38

(12  1)(30.0833)
14
2 = = 23.64

Since 2 = 23.64 > 2.025,11 = 21.9200, the decision is to reject the

null

hypothesis.

9.33 H0: 2 = .001  = .01 n = 16 df = 16 – 1 = 15 s2 = .

00144667

Ha: 2 > .001

2.01,15 = 30.5780

(16  1)(. 00144667)

.001
2 = = 21.7

Since 2 = 21.7 < 2.01,15 = 30.5780, the decision is to fail to reject

the null

hypothesis.
 P a g e | 39

9.34 H0: 2 = 199,996,164  = .10 /2 = .05 n = 13 df =13

- 1 = 12

Ha: 2  199,996,164 s2 = 832,089,743.7

2.05,12 = 21.0261 2.95,12 = 5.22603

(13  1)(832,089,743.6)
199,996,164
2 = = 49.93

Since 2 = 49.93 > 2.05,12 = 21.0261, the decision is to reject the

null

hypothesis. The variance has changed.

9.35 H0: 2 = .04  = .01 n=7 df = 7 – 1 = 6 s = .34 s2

= .1156

Ha: 2 > .04

2.01,6 = 16.8119

(7  1)(.1156)
.04
2 = = 17.34

Since 2 = 17.34 > 2.01,6 = 16.8119, the decision is to reject the

null hypothesis
 P a g e | 40

9.36 H0: µ = 100

Ha: µ < 100

n = 48 µ = 99  = 14

a)  = .10 z.10 = -1.28

xc  

n
zc =

x c  100
14
48
-1.28 =

x
c = 97.4

xc   97.4  99
 14
n 48
z = = = -0.79

from Table A.5, area for z = -0.79 is .2852

 P a g e | 41

 = .2852 + .5000 = .7852

b)  = .05 z.05 = -1.645

xc  

n
zc =

x c  100
14
48
-1.645 =

x
c = 96.68

xc   96.68  99
 14
n 48
z = = = -1.15

from Table A.5, area for z = -1.15 is .3749

 = .3749 + .5000 = .8749

 P a g e | 42

c)  = .01 z.01 = -2.33

xc  

n
zc =

x c  100
14
48
-2.33 =

x
c = 95.29

xc   95.29  99
 14
n 48
z = = = -1.84

from Table A.5, area for z = -1.84 is .4671

 = .4671 + .5000 = .9671

d) As  gets smaller (other variables remaining constant), gets

larger. Decreasing the probability of committing a Type I error
increases the probability of committing a Type II error if other
variables are held constant.
 P a g e | 43

9.37  = .05 µ = 100 n = 48  = 14

a) µa = 98.5 zc = -1.645

xc  

n
zc =

x c  100
14
48
-1.645 =

x
c = 96.68

xc   96.68  98.5
 14
n 48
z = = = -0.90

from Table A.5, area for z = -0.90 is .3159

 P a g e | 44

 = .3159 + .5000 = .8159

 P a g e | 45

b) µa = 98 zc = -1.645

x
c = 96.68

xc   96.68  98
 14
n 48
zc = = = -0.65

from Table A.5, area for z = -0.65 is .2422

 = .2422 + .5000 = .7422

c) µa = 97 z.05 = -1.645

x
c = 96.68

xc   96.68  97
 14
n 48
z = = = -0.16

from Table A.5, area for z = -0.16 is .0636

 = .0636 + .5000 = .5636

 P a g e | 46

d) µa = 96 z.05 = -1.645

x
c = 96.68

xc   96.68  96
 14
n 48
z = = = 0.34

from Table A.5, area for z = 0.34 is .1331

 = .5000 - .1331 = .3669

e) As the alternative value gets farther from the null hypothesized value, the

probability of committing a Type II error reduces (all other variables being held

constant).

9.38 Ho: µ = 50

Ha: µ  50

µa = 53 n = 35  =7  = .01

Since this is two-tailed, /2 = .005 z.005 = ±2.575

 P a g e | 47

xc  

n
zc =

x c  50
7
35
±2.575 =

x
c = 50 ± 3.05

46.95 and 53.05

xc   53.05  53
 7
n 35
z = = = 0.04

from Table A.5 for z = 0.04, area = .0160

Other end:

xc   46.95  53
 7
n 35
z = = = -5.11
 P a g e | 48

Area associated with z = -5.11 is .5000

 = .5000 + .0160 = .5160

 P a g e | 49

9.39 a) Ho: p = .65

Ha: p < .65

n = 360  = .05 pa = .60 z.05 = -1.645

pˆ c  p
pq
n
zc =

pˆ c  .65
(.65)(.35)
360
-1.645 =

p̂
c = .65 - .041 = .609

pˆ c  p .609  .60
pq (.60)(. 40)
n 360
z = = = 0.35

from Table A.5, area for z = -0.35 is .1368

 = .5000 - .1368 = .3632

p̂
b) pa = .55 z.05 = -1.645 c = .609
 P a g e | 50

pˆ c  P .609  .55
pq (.55)(.45)
n 360
z = = = 2.25

from Table A.5, area for z = -2.25 is .4878

 = .5000 - .4878 = .0122

p̂
c) pa = .50 z.05 = -1.645 c = .609

pˆ c  p .609  .50
pq (.50)(. 50)
n 360
z = = = -4.14

from Table A.5, the area for z = -4.14 is .5000

 = .5000 - .5000 = .0000

x
9.40 n = 58 = 45.1  = 8.7  = .05 /2 = .
025

H0: µ = 44

Ha: µ  44 z.025 = ± 1.96

 P a g e | 51

45.1  44
8 .7
58
z = = 0.96

Since z = 0.96 < zc = 1.96, the decision is to fail to reject the null
hypothesis.

x c  44
8.7
58
+ 1.96 =

x
± 2.239 = c - 44

x
c = 46.239 and 41.761

For 45 years:

46.239  45
8.7
58
z = = 1.08

from Table A.5, the area for z = 1.08 is .3599

 P a g e | 52

 = .5000 + .3599 = .8599

Power = 1 -  = 1 - .8599 = .1401

For 46 years:

46.239  46
8.7
58
z = = 0.21

From Table A.5, the area for z = 0.21 is .0832

 = .5000 + .0832 = .5832

Power = 1 -  = 1 - .5832 = .4168

For 47 years:
 P a g e | 53

46.239  47
8.7
58
z = = -0.67

From Table A.5, the area for z = -0.67 is .2486

 = .5000 - .2486 = .2514

Power = 1 -  = 1 - .2514 = .7486

For 48 years:

46.239  48
8.7
58
z = = 1.54

From Table A.5, the area for z = 1.54 is .4382

 = .5000 - .4382 = .0618

Power = 1 -  = 1 - .0618 = .9382

 P a g e | 54

9.41 H0: p = .71

Ha: p < .71

324
p̂ 463
n = 463 x = 324 = = .6998  = .10
 P a g e | 55

z.10 = -1.28

pˆ  p .6998  .71

pq (.71)(.29)
n 463
z = = -0.48

Since the observed z = -0.48 > z.10 = -1.28, the decision is to fail to
reject the null hypothesis.

Type II error:

p̂
Solving for the critical proportion, :
c

pˆ c  p
pq
n
zc =

pˆ c  .71
(.71)(. 29)
463
-1.28 =

p̂
= .683
 P a g e | 56

For pa = .69

.683  .69
(.69)(.31)
463
z = = -0.33

From Table A.5, the area for z = -0.33 is .1293

The probability of committing a Type II error = .1293 + .5000 = .6293

For pa = .66

.683  .66
(.66)(. 34)
463
z = = 1.04

From Table A.5, the area for z = 1.04 is .3508

The probability of committing a Type II error = .5000 - .3508 = .1492

For pa = .60
 P a g e | 57

.683  .60
(.60)(. 40)
463
z = = 3.65

From Table A.5, the area for z = 3.65 is essentially, .5000

The probability of committing a Type II error = .5000 - .5000 = .0000

 P a g e | 58

9.42 HTAB steps:

1)Ho: µ = 36

Ha: µ  36

x

n
2)z =

3)  = .01

4) two-tailed test, /2 = .005, z.005 = + 2.575

If the observed value of z is greater than 2.575 or less than -2.575, the
decision will be to reject the null hypothesis.

x
5) n = 63, = 38.4,  = 5.93

x 38.4  36
 5.93
n 63
6) z = = = 3.21

7) Since the observed value of z = 3.21 is greater than z.005 = 2.575, the
decision is

to reject the null hypothesis.

8) The mean is likely to be greater than 36.

 P a g e | 59

9.43 HTAB steps:

1)Ho: µ = 7.82

Ha: µ < 7.82

2) The test statistic is

x
s
n
t =

3)  = .05

4) df = n - 1 = 16, t.05,16 = -1.746. If the observed value of t is less than -1.746, then
the decision will be to reject the null hypothesis.

x
5) n = 17 = 7.01 s = 1.69

x 7.01  7.82

s 1.69
n 17
6) t = = = -1.98

7) Since the observed t = -1.98 is less than the table value of t = -1.746,
the decision

is to reject the null hypothesis.

 P a g e | 60

8) The population mean is significantly less than 7.82.

9.44 HTAB steps:

a. 1) Ho: p = .28
Ha: p > .28

pˆ  p
pq
n
2) z =
3)  = .10

4) This is a one-tailed test, z.10 = 1.28. If the observed value of z is greater

than 1.28, the decision will be to reject the null hypothesis.

5) n = 783 x = 230

230
pˆ 
783
= .2937

.2937  .28
(.28)(. 72)
783
6) z = = 0.85

7) Since z = 0.85 is less than z.10 = 1.28, the decision is to fail to

reject the null hypothesis.

8) There is not enough evidence to declare that p > .28.

 P a g e | 61

b. 1) Ho: p = .61
Ha: p  .61

pˆ  p
pq
n
2) z =

3)  = .05

4) This is a two-tailed test, z.025 = + 1.96. If the observed value of z is greater

than 1.96 or less than -1.96, then the decision will be to reject the null
hypothesis.

p̂
5) n = 401 = .56

.56  .61
(.61)(.39)
401
6) z = = -2.05

7) Since z = -2.05 is less than z.025 = -1.96, the decision is to reject the null
hypothesis.

8) The population proportion is not likely to be .61.

9.45 HTAB steps:

1) H0: 2 = 15.4

Ha: 2 > 15.4

 P a g e | 62

( n  1) s 2
2
2) 2 =

3)  = .01

4) n = 18, df = 17, one-tailed test

2.01,17 = 33.4087

5) s2 = 29.6

(n  1) s 2 (17)( 29.6)
2 15.4
6) 2 = = = 32.675

7) Since the observed 2 = 32.675 is less than 33.4087, the

decision is to fail

to reject the null hypothesis.

8) The population variance is not significantly more than 15.4.

9.46 a) H0: µ = 130

Ha: µ > 130

n = 75  = 12  = .01 z.01 = 2.33 µa = 135

 P a g e | 63

x
Solving for c:

xc  

n
zc =

x c  130
12
75
2.33 =

x
c = 133.23

133.23  135
12
75
z = = -1.28

from table A.5, area for z = -1.28 is .3997

 = .5000 - .3997 = .1003

b) H0: p = .44
Ha: p < .44
 P a g e | 64

n = 1095  = .05 pa = .42 z.05 = -1.645

pˆ c  p
pq
n
zc =

pˆ c  .44
(.44)(.56)
1095
-1.645 =

p̂c
= .4153

.4153  .42
(.42)(. 58)
1095
z = = -0.32

from table A.5, area for z = -0.32 is .1255

 = .5000 + .1255 = .6255

9.47 H0: p = .32

Ha: p > .32

p̂
n = 80  = .01 = .39
 P a g e | 65

z.01 = 2.33

pˆ  p .39  .32

pq (.32)(.68)
n 80
z = = 1.34

Since the observed z = 1.34 < z.01 = 2.33, the decision is to fail to
reject the null hypothesis.

x
9.48 = 3.45 n = 64 2 = 1.31  = .05

Ho: µ = 3.3

Ha: µ  3.3

For two-tail, /2 = .025 zc = ±1.96

x 3.45  3.3

 1.31
n 64
z = = = 1.05

Since the observed z = 1.05 < zc = 1.96, the decision is to Fail to

reject the null hypothesis.
 P a g e | 66

x 93
pˆ  
n 210
9.49 n = 210 x = 93  = .10 = .
443

Ho: p = .57

Ha: p < .57

For one-tail,  = .10 zc = -1.28

pˆ  p .443  .57

pq (.57)(.43)
n 210
z = = -3.72

Since the observed z = -3.72 < zc = -1.28, the decision is to reject the
null hypothesis.
 P a g e | 67

9.50 H0: 2 = 16 n = 12  = .05 df = 12 - 1 = 11

Ha: 2 > 16

s = 0.4987864 ft. = 5.98544 in.

2.05,11 = 19.6752

(12  1)(5.98544) 2
16
2 = = 24.63

Since 2 = 24.63 > 2.05,11 = 19.6752, the decision is to reject the

null hypothesis.

9.51 H 0: µ = 8.4  = .01 /2 = .005 n=7 df = 7 –

1=6 s = 1.3

Ha: µ  8.4

x
= 5.6 t.005,6 = + 3.707

5.6  8.4
1. 3
7
t = = -5.70
 P a g e | 68

Since the observed t = - 5.70 < t.005,6 = -3.707, the decision is to reject
the null hypothesis.

x
9.52 = $26,650 n = 100  = $12,000

a) Ho: µ = $25,000

Ha: µ > $25,000  = .05

For one-tail,  = .05 z.05 = 1.645

x 26,650  25,000

 12,000
n 100
z = = = 1.38

Since the observed z = 1.38 < z.05 = 1.645, the decision is to fail
to reject the null hypothesis.

b) µa = $30,000 zc = 1.645

x
Solving for :
c
 P a g e | 69

xc  

n
zc =

( x c  25,000)
12,000
100
1.645 =

x
c = 25,000 + 1,974 = 26,974

26,974  30,000
12,000
100
z = = -2.52

from Table A.5, the area for z = -2.52 is .4941

 = .5000 - .4941 = .0059

 P a g e | 70

9.53 H0: 2 = 4 n=8 s = 7.80  = .10 df = 8 – 1

=7

Ha: 2 > 4

2.10,7 = 12.0170

(8  1)(7.80) 2
4
2 = = 106.47

Since observed 2 = 106.47 > 2.10,7 = 12.017, the decision is to

reject the null hypothesis.

9.54 H0: p = .46

Ha: p > .46

x 66
pˆ  
n 125
n = 125 x = 66  = .05 = .528

Using a one-tailed test, z.05 = 1.645

pˆ  p .528  .46

pq (.46)(.54)
n 125
z = = 1.53

Since the observed value of z = 1.53 < z.05 = 1.645, the decision is to
fail to reject the null hypothesis.
 P a g e | 71

p̂c
Solving for :

pˆ c  p
pq
n
zc =

pˆ c  .46
(.46)(. 54)
125 p̂c
1.645 = and therefore, = .533

pˆ c  p a .533  .50

pa  qa (.50)(. 50)
n 125
z = = 0.74

from Table A.5, the area for z = 0.74 is .2704

 = .5000 + .2704 = .7704

x
9.55 n = 16 = 175 s = 14.28286 df = 16 - 1 = 15 =.
05

H0: µ = 185

Ha: µ < 185

t.05,15 = - 1.753
 P a g e | 72

x 175  185

s 14.28286
n 16
t = = = -2.80

Since observed t = - 2.80 < t.05,15 = - 1.753, the decision is to reject

the null hypothesis.

9.56 H0: p = .182

Ha: p > .182

x 84
pˆ  
n 428
n = 428 x = 84  = .01 = .1963

For a one-tailed test, z.01 = 2.33

pˆ  p .1963  .182

pq (.182)(.818)
n 428
z = = 0.77

Since the observed z = 0.77 < z.01 = 2.33, the decision is to fail to
reject the null hypothesis.

The probability of committing a Type I error is .01.

 P a g e | 73

p̂c
Solving for :

pˆ c  p
pq
n
zc =

. pˆ c  .182
(.182)(.818)
428
2.33 =

p̂c
= .2255

pˆ c  p a .2255  .21

pa  qa (.21)(.79)
n 428
z = = 0.79

from Table A.5, the area for z = 0.79 is .2852

 = .5000 + .2852 = .7852

9.57 Ho: µ = $15

Ha: µ > $15

 P a g e | 74

x
= $19.34 n = 35  = $4.52  = .10

For one-tail and  = .10 zc = 1.28

x 19.34  15
s 4.52
n 35
z = = = 5.68

Since the observed z = 5.68 > zc = 1.28, the decision is to reject the
null hypothesis.

9.58 H0: 2 = 16 n = 22 df = 22 –1 = 21 s=6  = .05

Ha: 2 > 16

2.05,21 = 32.6706

(22  1)(6) 2
16
2 = = 47.25

Since the observed 2 = 47.25 > 2.05,21 = 32.6706, the decision is to

reject the null hypothesis.
 P a g e | 75

x
9.59 H 0: µ = 2.5 = 3.4 s = 0.6  = .01 n=9 df =
9–1=8

Ha: µ > 2.5

t.01,8 = 2.896

x 3.4  2.5

s 0.6
n 9
t = = = 4.50

Since the observed t = 4.50 > t.01,8 = 2.896, the decision is to reject
the null hypothesis.

9.60 a) Ho:  = 23.58

Ha:   23.58

x
n = 95 = 22.83  = 5.11  = .05

Since this is a two-tailed test and using /2 = .025: z.025 = +

1.96

x 22.83  23.58

 5.11
n 95
z = = = -1.43
 P a g e | 76

Since the observed z = -1.43 > z.025 = -1.96, the decision is to

fail to reject the null hypothesis.

xc  

n
b) zc =

xc  23.58
5.11
95
+ 1.96 =

xc
= 23.58 + 1.03

xc
= 22.55, 24.61

for Ha:  = 22.30

xc   a 22.55  22.30
 5.11
n 95
z = = = 0.48
 P a g e | 77

xc   a 24.61  22.30
 5.11
n 95
z = = = 4.41

from Table A.5, the areas for z = 0.48 and z = 4.41 are .1844
and .5000

 = .5000 - .1844 = .3156

The upper tail has no effect on .

x
9.61 n = 12 = 12.333 s2 = 10.424

H0: 2 = 2.5

Ha: 2  2.5

 = .05 df = 11 two-tailed test, /2 = .025

2.025,11 = 21.9200

2..975,11 = 3.81574
 P a g e | 78

If the observed 2 is greater than 21.9200 or less than 3.81574, the

decision is to reject the null hypothesis.

( n  1) s 2 11(10.424)
2 2 .5
2 = = = 45.866

Since the observed 2 = 45.866 is greater than 2.025,11 = 21.92, the

decision is to reject the null hypothesis. The population variance is
significantly more than 2.5.
 P a g e | 79

x
9.62 H0: µ = 23 = 18.5 s = 6.91  = .10 n = 16 df =
16 – 1 = 15

Ha: µ < 23

t.10,15 = -1.341

x 18.5  23
s 6.91
n 16
t = = = -2.60

Since the observed t = -2.60 < t.10,15 = -1.341, the decision is to reject
the null hypothesis.

x
9.63 The sample size is 22. is 3.969 s = 0.866 df = 21

The test statistic is:

x
s
n
t =

The observed t = -2.33. The p-value is .015.

The results are statistical significant at  = .05.

 P a g e | 80

The decision is to reject the null hypothesis.

9.64 H0: p = .25

Ha: p  .25

This is a two-tailed test with  = .05. n = 384.

Since the p-value = .045 <  = .05, the decision is to reject the null
hypothesis.

p̂
The sample proportion, = .205729 which is less than the
hypothesized p = .25.

One conclusion is that the population proportion is lower than .25.

 P a g e | 81

9.65 H0:  = 2.51

Ha:  > 2.51

This is a one-tailed test. The sample mean is 2.55 which is more than
the hypothesized value. The observed t value is 1.51 with an
associated

p-value of .072 for a one-tailed test. Because the p-value is greater

than

 = .05, the decision is to fail to reject the null hypothesis.

There is not enough evidence to conclude that beef prices are higher.

9.66 H0:  = 2747

Ha:  < 2747

This is a one-tailed test. Sixty-seven households were included in this

study.

The sample average amount spent on home-improvement projects was

2,349.

Since z = -2.09 < z.05 = -1.645, the decision is to reject the null
hypothesis at

 = .05. This is underscored by the p-value of .018 which is less than 

= .05.

However, the p-value of .018 also indicates that we would not reject
the null

hypothesis at  = .01.
 P a g e | 82

Chapter 10
Statistical Inferences about Two Populations

LEARNING OBJECTIVES

The general focus of Chapter 10 is on testing hypotheses and constructing

confidence intervals about parameters from two populations, thereby
enabling you to

1. Test hypotheses and construct confidence intervals about the

difference in two population means using the z statistic.

2. Test hypotheses and establish confidence intervals about the

difference in two population means using the t statistic.

3. Test hypotheses and construct confidence intervals about the

difference in two related populations.

4. Test hypotheses and construct confidence intervals about the

difference in two population proportions.

5. Test hypotheses and construct confidence intervals about two

population variances.
 P a g e | 83

CHAPTER TEACHING STRATEGY

The major emphasis of chapter 10 is on analyzing data from two

samples. The student should be ready to deal with this topic given that
he/she has tested hypotheses and computed confidence intervals in previous
chapters on single sample data.

In this chapter, the approach as to whether to use a z statistic or a t

statistic for analyzing the differences in two sample means is the same as
that used in chapters 8

and 9. When the population variances are known, the z statistic can be used.
However, if the population variances are unknown and sample variances are
being used, then the

t test is the appropriate statistic for the analysis. It is always an assumption

underlying the use of the t statistic that the populations are normally
distributed. If sample sizes are small and the population variances are
known, the z statistic can be used if the populations are normally distributed.
 P a g e | 84

In conducting a t test for the difference of two means from

independent populations, there are two different formulas given in the
chapter. One version of this test uses a "pooled" estimate of the population
variance and assumes that the population variances are equal. The other
version does not assume equal population variances and is simpler to
compute. In doing hand calculations, it is generally easier to use the
“pooled” variance formula because the degrees of freedom formula for the
unequal variance formula is quite complex. However, it is good to expose
students to both formulas since computer software packages often give you
the option of using the “pooled” that assumes equal population variances or
the formula for unequal variances.

A t test is also included for related (non independent) samples. It is

important that the student be able to recognize when two samples are
related and when they are independent. The first portion of section 10.3
addresses this issue. To underscore the potential difference in the outcome
of the two techniques, it is sometimes valuable to analyze some related
measures data with both techniques and demonstrate that the results and
conclusions are usually quite different. You can have your students work
problems like this using both techniques to help them understand the
differences between the two tests (independent and dependent t tests) and
the different outcomes they will obtain.

A z test of proportions for two samples is presented here along with an

F test for two population variances. This is a good place to introduce the
student to the F distribution in preparation for analysis of variance in Chapter
11. The student will begin to understand that the F values have two different
degrees of freedom. The F distribution tables are upper tailed only. For this
reason, formula 10.14 is given in the chapter to be used to compute lower
tailed F values for two-tailed tests.
 P a g e | 85

CHAPTER OUTLINE

10.1 Hypothesis Testing and Confidence Intervals about the Difference in

Two Means using the z Statistic (Population Variances Known)

Hypothesis Testing

Confidence Intervals

Using the Computer to Test Hypotheses about the Difference in

Two

Population Means Using the z Test

10.2 Hypothesis Testing and Confidence Intervals about the Difference in

Two Means: Independent Samples and Population Variances Unknown

Hypothesis Testing

Using the Computer to Test Hypotheses and Construct

Confidence

Intervals about the Difference in Two Population Means Using

the t

Test

Confidence Intervals

10.3 Statistical Inferences For Two Related Populations

Hypothesis Testing

Using the Computer to Make Statistical Inferences about Two

Related

Populations

Confidence Intervals
 P a g e | 86

10.4 Statistical Inferences About Two Population Proportions, p1- p2

Hypothesis Testing

Confidence Intervals

Using the Computer to Analyze the Difference in Two Proportions

10.5 Testing Hypotheses About Two Population Variances

Using the Computer to Test Hypotheses about Two Population

Variances

KEY TERMS

Dependent Samples Independent

Samples

F Distribution Matched-Pairs Test

F Value Related Measures

 P a g e | 87

SOLUTIONS TO PROBLEMS IN CHAPTER 10

10.1 Sample 1 Sample 2

x x
1 = 51.3 2 = 53.2

s12 = 52 s22 = 60

n1 = 31 n2 = 32

a) Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 < 0

For one-tail test,  = .10 z.10 = -1.28

( x 1  x 2 )  ( 1   2 ) (51.3  53.2)  (0)


1  2
2 2
52 60
 
n1 n2 31 32
z = = -1.01

Since the observed z = -1.01 > zc = -1.28, the decision is to fail

to reject the null hypothesis.

b) Critical value method:

 P a g e | 88

( x 1  x 2 ) c  ( 1   2 )
1  2
2 2


n1 n2
zc =

( x 1  x 2 ) c  ( 0)
52 60

31 32
-1.28 =

x x
( 1- 2)c = -2.41

c) The area for z = -1.01 using Table A.5 is .3438.

The p-value is .5000 - .3438 = .1562

 P a g e | 89

10.2 Sample 1 Sample 2

n1 = 32 n2 = 31

x x
1 = 70.4 2 = 68.7

1 = 5.76 2 = 6.1

For a 90% C.I., z.05 = 1.645

 
2 2

( x1  x 2 )  z 1  2
n1 n2

5.76 2 6.12

32 31
(70.4) – 68.7) + 1.645

1.7 ± 2.46

-.76 < µ1 - µ2 < 4.16

10.3 a) Sample 1 Sample 2

x x
1 = 88.23 2 = 81.2
 P a g e | 90

12 = 22.74 22 = 26.65

n1 = 30 n2 = 30

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

For two-tail test, use /2 = .01 z.01 = + 2.33

( x 1  x 2 )  ( 1   2 ) (88.23  81.2)  (0)


1  2
2 2
22.74 26.65
 
n1 n2 30 30
z = = 5.48

Since the observed z = 5.48 > z.01 = 2.33, the decision is to

reject the null hypothesis.
 P a g e | 91

1  2
2 2

( x1  x 2 )  z 
n1 n2
b)

22.74 26.65

30 30
(88.23 – 81.2) + 2.33

7.03 + 2.99

4.04 <  < 10.02

This supports the decision made in a) to reject the null hypothesis because zero is not in
the interval.

10.4 Computers/electronics Food/Beverage

x x
1 = 1.96 2 = 3.02

12 = 1.0188 22 = .9180

n1 = 50 n2 = 50

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

For two-tail test, /2 = .005 z.005 = ±2.575

 P a g e | 92

( x 1  x 2 )  ( 1   2 ) (1.96  3.02)  (0)


1  2
2 2
1.0188 0.9180
 
n1 n2 50 50
z = = -5.39

Since the observed z = -5.39 < zc = -2.575, the decision is to reject

the null hypothesis.
 P a g e | 93

10.5 A B

n1 = 40 n2 = 37

x x
1 = 5.3 2 = 6.5

12 = 1.99 22 = 2.36

For a 95% C.I., z.025 = 1.96

1  2
2 2

( x1  x 2 )  z 
n1 n2

1.99 2.36

40 37
(5.3 – 6.5) + 1.96

-1.2 ± .66 -1.86 <  < -.54

The results indicate that we are 95% confident that, on average,

Plumber B does between 0.54 and 1.86 more jobs per day than
Plumber A. Since zero does not lie in this interval, we are confident
that there is a difference between Plumber A and Plumber B.

10.6 Managers Specialty

n1 = 35 n2 = 41
 P a g e | 94

x x
1 = 1.84 2 = 1.99

1 = .38 2 = .51

for a 98% C.I., z.01 = 2.33

 
2 2

( x1  x 2 )  z 1  2
n1 n2

.382 .512

35 41
(1.84 - 1.99) ± 2.33

-.15 ± .2384

-.3884 < µ1 - µ2 < .0884

Point Estimate = -.15

Hypothesis Test:

1) Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0
 P a g e | 95

( x 1  x 2 )  ( 1   2 )
1  2
2 2


n1 n2
2) z =

3)  = .02

4) For a two-tailed test, z.01 = + 2.33. If the observed z value is greater

than 2.33

or less than -2.33, then the decision will be to reject the null
hypothesis.

5) Data given above

(1.84  1.99)  (0)

(.38) 2 (.51) 2

35 41
6) z = = -1.47

7) Since z = -1.47 > z.01 = -2.33, the decision is to fail to reject the
null

hypothesis.

8) There is no significant difference in the hourly rates of the two

groups.

10.7 1996 2006

 P a g e | 96

x x
1 = 190 2 = 198

1 = 18.50 2 = 15.60

n1 = 51 n2 = 47  = .01

H0: 1 - 2 = 0

Ha: 1 - 2 < 0

For a one-tailed test, z.01 = -2.33

( x 1  x 2 )  ( 1   2 ) (190  198)  (0)


1  2
2 2
(18.50) 2 (15.60) 2
 
n1 n2 51 47
z = = -2.32

Since the observed z = -2.32 > z.01 = -2.33, the decision is to fail to
reject the null hypothesis.

10.8 Seattle Atlanta

n1 = 31 n2 = 31

x x
1 = 2.64 2 = 2.36

12 = .03 22 = .015

For a 99% C.I., z.005 = 2.575

 P a g e | 97

1  2
2 2

( x1  x 2 )  z 
n1 n2

.03 .015

31 31
(2.64-2.36) ± 2.575

.28 ± .10 .18 <  < .38

Between $ .18 and $ .38 difference with Seattle being more expensive.

10.9 Canon Pioneer

x x
1 = 5.8 2 = 5.0

1 = 1.7 2 = 1.4

n1 = 36 n2 = 45

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

For two-tail test, /2 = .025 z.025 = ±1.96

 P a g e | 98

( x 1  x 2 )  ( 1   2 ) (5.8  5.0)  (0)


1  2
2 2
(1.7) 2 (1.4)
 
n1 n2 36 45
z = = 2.27

Since the observed z = 2.27 > zc = 1.96, the decision is to reject the
null

hypothesis.
 P a g e | 99

10.10 A B

x x
1 = 8.05 2 = 7.26

1 = 1.36 2 = 1.06

n1 = 50 n2 = 38

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 > 0

For one-tail test,  = .10 z.10 = 1.28

( x 1  x 2 )  ( 1   2 ) (8.05  7.26)  (0)


1  2
2 2
(1.36) 2 (1.06) 2
 
n1 n2 50 38
z = = 3.06

Since the observed z = 3.06 > zc = 1.28, the decision is to reject the
null hypothesis.

10.11 Ho: µ1 - µ2 = 0  = .01

Ha: µ1 - µ2 < 0 df = 8 + 11 - 2 = 17

Sample 1 Sample 2

n1 = 8 n2 = 11
 P a g e | 100

x x
1 = 24.56 2 = 26.42

s12 = 12.4 s22 = 15.8

For one-tail test,  = .01 Critical t.01,17 = -2.567

( x1  x 2 )  ( 1   2 )
(24.56  26.42)  (0)
2 2
s1 ( n1  1)  s 2 (n 2  1) 1 1 12.4(7)  15.8(10) 1 1
 
n1  n 2  2 n1 n2 8  11  2 8 11
t = = =
-1.05

Since the observed t = -1.05 > t.01,19 = -2.567, the decision is to fail to
reject the null hypothesis.
 P a g e | 101

10.12 a) Ho: µ1 - µ2 = 0  =.10

Ha: µ1 - µ2  0 df = 20 + 20 - 2 = 38

Sample 1 Sample 2

n1 = 20 n2 = 20

x x
1 = 118 2 = 113

s1 = 23.9 s2 = 21.6

For two-tail test, /2 = .05 Critical t.05,38 = 1.697 (used

df=30)

( x1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n 2  1) 1 1

n1  n2  2 n1 n2
t = =

(118  113)  (0)

(23.9) (19)  (21.6) 2 (19) 1
2
1

20  20  2 20 20
t = = 0.69

Since the observed t = 0.69 < t.05,38 = 1.697, the decision is to

fail to reject the null hypothesis.
 P a g e | 102

2 2
s1 (n1  1)  s 2 (n 2  1) 1 1
( x1  x 2 )  t 
n1  n2  2 n1 n 2
b) =

(23.9) 2 (19)  (21.6) 2 (19) 1 1


20  20  2 20 20
(118 – 113) + 1.697

5 + 12.224

-7.224 <  1 -  2 < 17.224

 P a g e | 103

10.13 Ho: µ1 - µ2 = 0  = .05

Ha: µ1 - µ2 > 0 df = n1 + n2 - 2 = 10 + 10 - 2 = 18

Sample 1 Sample 2

n1 = 10 n2 = 10

x x
1 = 45.38 2 = 40.49

s1 = 2.357 s2 = 2.355

For one-tail test,  = .05 Critical t.05,18 = 1.734

( x1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n 2  1) 1 1

n1  n2  2 n1 n2
t = =

(45.38  40.49)  (0)

(2.357) 2 (9)  (2.355) 2 (9) 1 1

10  10  2 10 10
t = = 4.64

Since the observed t = 4.64 > t.05,18 = 1.734, the decision is to

reject the null hypothesis.

10.14 Ho: µ1 - µ2 = 0  =.01

 P a g e | 104

Ha: µ1 - µ2  0 df = 18 + 18 - 2 = 34

Sample 1 Sample 2

n1 = 18 n2 = 18

x x
1 = 5.333 2 = 9.444

s12 = 12 s22 = 2.026

For two-tail test, /2 = .005 Critical t.005,34 = ±2.75 (used

df=30)

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t = =

(5.333  9.444)  (0)

12(17)  (2.026)17 1 1

18  18  2 18 18
t = = -4.66

Since the observed t = -4.66 < t.005,34 = -2.75, reject the null
hypothesis.

b) For 98% confidence, t.01, 30 = 2.457

 P a g e | 105

2 2
s1 (n1  1)  s 2 (n 2  1) 1 1
( x1  x 2 )  t 
n1  n2  2 n1 n 2
=

(12)(17)  (2.026)(17) 1 1

18  18  2 18 18
(5.333 – 9.444) + 2.457

-4.111 + 2.1689

-6.2799 <  1 -  2 < -1.9421

10.15 Peoria Evansville

n1 = 21 n2 = 26

x1 x2
= 116,900 = 114,000

s1 = 2,300 s2 = 1,750 df = 21 + 26 – 2

90% level of confidence, /2 = .05 t .05,45 = 1.684 (used

df = 40)

2 2
s (n  1)  s 2 (n2  1) 1 1
( x1  x 2 )  t 1 1 
n1  n 2  2 n1 n2
=
 P a g e | 106

(2300) 2 (20)  (1750) 2 (25) 1 1


21  26  2 21 26
(116,900 – 114,000) + 1.684
=

2,900 + 994.62

1905.38 <  1 -  2 < 3894.62

 P a g e | 107

10.16 Ho: µ1 - µ2 = 0  = .10

Ha: µ1 - µ2  0 df = 12 + 12 - 2 = 22

Co-op Interns

n1 = 12 n2 = 12

x x
1 = $15.645 2 = $15.439

s1 = $1.093 s2 = $0.958

For two-tail test, /2 = .05

Critical t.05,22 = ± 1.717

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t = =

(15.645  15.439)  (0)

(1.093) 2 (11)  (0.958) 2 (11) 1 1

12  12  2 12 12
t = = 0.49

Since the observed t = 0.49 < t.05,22 = 1.717, the decision is to fail
reject the null hypothesis.
 P a g e | 108

90% Confidence Interval: t.05,22 = ± 1.717

2 2
s (n  1)  s 2 (n2  1) 1 1
( x1  x 2 )  t 1 1 
n1  n 2  2 n1 n2
=

(1.093) 2 (11)  (0.958) 2 (11) 1 1


12  12  2 12 12
(15.645 – 15.439) + 1.717 =

0.206 + 0.7204

-0.5144 <  1 -  2 < 0.9264

 P a g e | 109

10.17 Let Boston be group 1

1) Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 > 0

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
2) t =

3)  = .01

4) For a one-tailed test and df = 8 + 9 - 2 = 15, t.01,15 = 2.602. If the

observed

value of t is greater than 2.602, the decision is to reject the null

hypothesis.

5) Boston Dallas

n1 = 8 n2 = 9

x x
1 = 47 2 = 44

s1 = 3 s2 = 3

(47  44)  (0)

7(3) 2  8(3) 2 1 1

15 8 9
6) t = = 2.06

7) Since t = 2.06 < t.01,15 = 2.602, the decision is to fail to reject the
null
 P a g e | 110

hypothesis.

8) There is no significant difference in rental rates between Boston

and Dallas.

10.18 nm = 22 nno = 20

x x
m = 112 no = 122

sm = 11 sno = 12

df = nm + nno - 2 = 22 + 20 - 2 = 40

For a 98% Confidence Interval, /2 = .01 and t.01,40 = 2.423

2 2
s (n  1)  s 2 (n2  1) 1 1
( x1  x 2 )  t 1 1 
n1  n 2  2 n1 n2
=

(11) 2 (21)  (12) 2 (19) 1 1


22  20  2 22 20
(112 – 122) + 2.423

-10 ± 8.60

-$18.60 < µ1 - µ2 < -$1.40

 P a g e | 111

Point Estimate = -$10

10.19 Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

df = n1 + n2 - 2 = 11 + 11 - 2 = 20

Toronto Mexico City

n1 = 11 n2 = 11

x x
1 = $67,381.82 2 = $63,481.82

s1 = $2,067.28 s2 = $1,594.25

For a two-tail test, /2 = .005 Critical t.005,20 = ±2.845

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t = =

(67,381.82  63,481.82)  (0)

(2,067.28) 2 (10)  (1,594.25) 2 (10) 1 1

11  11  2 11 11
t = = 4.95
 P a g e | 112

Since the observed t = 4.95 > t.005,20 = 2.845, the decision is to

Reject the null

hypothesis.

10.20 Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 > 0

df = n1 + n2 - 2 = 9 + 10 - 2 = 17

Men Women

n1 = 9 n2 = 10

x x
1 = $110.92 2 = $75.48

s1 = $28.79 s2 = $30.51

This is a one-tail test,  = .01 Critical t.01,17 = 2.567

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t = =

( 1 1 0 .9 2  7 5 .4 8 )  ( 0 )
( 2 8 . 7 9 ) 2 ( 8 )  ( 3 0 .5 1 ) 2 ( 9 ) 1 1

9  10  2 9 10
t = = 2.60
 P a g e | 113

Since the observed t = 2.60 > t.01,17 = 2.567, the decision is to Reject
the null

hypothesis.

10.21 Ho: D = 0

Ha: D > 0

Sample 1 Sample 2 d

38 22 16

27 28 -1

30 21 9

41 38 3

36 38 -2

38 26 12

33 19 14

35 31 4

44 35 9

d
n=9 =7.11 sd=6.45  = .01

df = n - 1 = 9 - 1 = 8
 P a g e | 114

For one-tail test and  = .01, the critical t.01,8 = ±2.896

d  D 7.11  0

sd 6.45
n 9
t = = 3.31

Since the observed t = 3.31 > t.01,8 = 2.896, the decision is to reject
the null

hypothesis.
 P a g e | 115

10.22 Ho: D=0

Ha: D0

Before After d

107 102 5

99 98 1

110 100 10
113 108 5

96 89 7

98 101 -3

100 99 1

102 102 0

107 105 2

109 110 -1

104 102 2

99 96 3

101 100 1

d
n = 13 = 2.5385 sd=3.4789  = .05

df = n - 1 = 13 - 1 = 12

For a two-tail test and /2 = .025 Critical t.025,12 = ±2.179

 P a g e | 116

d  D 2.5385  0

sd 3.4789
n 13
t = = 2.63

Since the observed t = 2.63 > t.025,12 = 2.179, the decision is to reject
the null

hypothesis.
 P a g e | 117

d
10.23 n = 22 = 40.56 sd = 26.58

For a 98% Level of Confidence, /2 = .01, and df = n - 1 = 22 - 1 = 21

t.01,21 = 2.518

sd
d t
n

26.58
22
40.56 ± (2.518)

40.56 ± 14.27

26.29 < D < 54.83

10.24 Before After d

32 40 -8

28 25 3

35 36 -1

32 32 0

26 29 -3
 P a g e | 118

25 31 -6

37 39 -2

16 30 -14

35 31 4

d
n=9 = -3 sd = 5.6347  = .025

df = n - 1 = 9 - 1 = 8

For 90% level of confidence and /2 = .05, t.05,8 = 1.86

sd
d t
n
t =

5.6347
9
t = -3 + (1.86) = -3 ± 3.49

-6.49 < D < 0.49

10.25 City Cost Resale d

Atlanta 2042725163 -4736

Boston 2725524625 2630

Des Moines 22115 12600 9515

 P a g e | 119

Kansas City 23256 24588 -1332

Louisville 21887 19267 2620

Portland 24255 20150 4105

Raleigh-Durham 19852 22500 -2648

Reno 23624 16667 6957

Ridgewood 25885 26875 - 990

San Francisco 28999 35333 -6334

Tulsa 20836 16292 4544

d
= 1302.82 sd = 4938.22 n = 11, df = 10

 = .01 /2 = .005 t.005,10= 3.169

sd 4938.22
d t
n 11
= 1302.82 + 3.169 = 1302.82 + 4718.42

-3415.6 < D < 6021.2

 P a g e | 120

10.26 Ho: D=0

Ha: D<0

Before After d

2 4 -2

4 5 -1

1 3 -2

3 3 0

4 3 1

2 5 -3

2 6 -4

3 4 -1

1 5 -4

d
n=9 =-1.778 sd=1.716  = .05 df =
n-1=9-1=8

For a one-tail test and  = .05, the critical t.05,8 = -1.86

d  D  1.778  0

sd 1.716
n 9
t = = -3.11
 P a g e | 121

Since the observed t = -3.11 < t.05,8 = -1.86, the decision is to reject
the null hypothesis.

10.27 Before After d

255 197 58

230 225 5

290 215 75

242 215 27

300 240 60

250 235 15

215 190 25

230 240 -10

225 200 25

219 203 16

236 223 13

d
n = 11 = 28.09 sd=25.813 df =
n - 1 = 11 - 1 = 10

For a 98% level of confidence and /2=.01, t.01,10 = 2.764

 P a g e | 122

sd
d t
n

25.813
11
28.09 ± (2.764) = 28.09 ± 21.51

6.58 < D < 49.60

 P a g e | 123

10.28 H0: D = 0

d
Ha: D > 0 n = 27 df = 27 – 1 = 26 = 3.17 sd = 5

Since  = .01, the critical t.01,26 = 2.479

d  D 3.71  0

sd 5
n 27
t = = 3.86

Since the observed t = 3.86 > t.01,26 = 2.479, the decision is to reject
the null hypothesis.

d
10.29 n = 21 = 75 sd = 30 df = 21 - 1 = 20

For a 90% confidence level, /2=.05 and t.05,20 = 1.725

sd
d t
n

30
21
75 + 1.725 = 75 ± 11.29
 P a g e | 124

63.71 < D < 86.29

10.30 Ho: D=0

Ha: D0

d
n = 15 = -2.85 sd = 1.9  = .01 df
= 15 - 1 = 14

For a two-tail test, /2 = .005 and the critical t.005,14 = + 2.977

d  D  2.85  0

sd 1.9
n 15
t = = -5.81

Since the observed t = -5.81 < t.005,14 = -2.977, the decision is to reject
the null

hypothesis.

10.31 a) Sample 1 Sample 2

n1 = 368 n2 = 405

x1 = 175 x2 = 182
 P a g e | 125

x1 175 x 2 182
pˆ 1   pˆ 2  
n1 368 n2 405
= .476 = .449

x1  x 2 175  182 357

p  
n1  n 2 368  405 773
= .462

Ho: p1 - p2 = 0

Ha: p1 - p2  0

For two-tail, /2 = .025 and z.025 = ±1.96

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.476  .449)  (0)

z 
 1 1  1 1 
p  q   (.462)(. 538)  
 n1 n   368 405 
 
= 0.75

Since the observed z = 0.75 < zc = 1.96, the decision is to fail to

reject the null

hypothesis.

b) Sample 1 Sample 2

p̂ p̂
1 = .38 2 = .25

n1 = 649 n2 = 558
 P a g e | 126

n1 pˆ 1  n2 pˆ 2 649(.38)  558(.25)
p 
n1  n 2 649  558
= .32

Ho: p1 - p2 = 0

Ha: p1 - p2 > 0

For a one-tail test and  = .10, z.10 = 1.28

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.38  .25)  (0)

z 
 1 1  1 1 
p  q   (.32)(. 68)  
 n1 n   649 558 
 
= 4.83

Since the observed z = 4.83 > zc = 1.28, the decision is to reject the
null

hypothesis.

p̂ p̂
10.32 a) n1 = 85 n2 = 90 1 = .75 2 = .67

For a 90% Confidence Level, z.05 = 1.645

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2
 P a g e | 127

(.75)(.25) (.67)(.33)

85 90
(.75 - .67) ± 1.645 = .08 ± .11

-.03 < p1 - p2 < .19

p̂ p̂
b) n1 = 1100 n2 = 1300 1 = .19 2 = .17

For a 95% Confidence Level, /2 = .025 and z.025 = 1.96

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.19)(. 81) (.17)(.83)


1100 1300
(.19 - .17) + 1.96 = .02 ± .03

-.01 < p1 - p2 < .05

c) n1 = 430 n2 = 399 x1 = 275 x2 = 275

x1 275 x 2 275
pˆ 1   pˆ 2  
n1 430 n2 399
= .64 = .69
 P a g e | 128

For an 85% Confidence Level, /2 = .075 and z.075 = 1.44

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.64)(. 36) (.69)(.31)


430 399
(.64 - .69) + 1.44 = -.05 ± .047

-.097 < p1 - p2 < -.003

 P a g e | 129

d) n1 = 1500 n2 = 1500 x1 = 1050 x2 = 1100

x1 1050 x 2 1100
pˆ 1   pˆ 2  
n1 1500 n 2 1500
= .70 = .733

For an 80% Confidence Level, /2 = .10 and z.10 = 1.28

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.70)(. 30) (.733)(.267)


1500 1500
(.70 - .733) ± 1.28 = -.033 ± .02

-.053 < p1 - p2 < -.013

10.33 H0: pm - pw = 0

p̂ p̂
Ha: pm - pw < 0 nm = 374 nw = 481
m = .59 w
= .70

For a one-tailed test and  = .05, z.05 = -1.645

 P a g e | 130

n m pˆ m  n w pˆ w 374(.59)  481(.70)
p 
nm  n w 374  481
= .652

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.59  .70)  (0)

z 
 1 1  1 1 
p  q   (.652)(.348)  
 n1 n   374 481 
 
= -3.35

Since the observed z = -3.35 < z.05 = -1.645, the decision is to reject
the null

hypothesis.
 P a g e | 131

p̂1 p̂2
10.34 n1 = 210 n2 = 176 = .24 = .35

For a 90% Confidence Level, /2 = .05 and z.05 = + 1.645

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.24)(. 76) (.35)(.65)


210 176
(.24 - .35) + 1.645 = -.11 + .0765

-.1865 < p1 – p2 < -.0335

10.35 Computer Firms Banks

p̂ p̂
1 = .48 2 = .56

n1 = 56 n2 = 89

n1 pˆ 1  n 2 pˆ 2 56(.48)  89(.56)
p 
n1  n2 56  89
= .529

Ho: p1 - p2 = 0
 P a g e | 132

Ha: p1 - p2  0

For two-tail test, /2 = .10 and zc = ±1.28

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.48  .56)  (0)

z 
 1 1  1 1 
p  q   (.529)(. 471)  
 n1 n   56 89 
 
= -0.94

Since the observed z = -0.94 > zc = -1.28, the decision is to fail to

reject the null

hypothesis.
 P a g e | 133

10.36 A B

n1 = 35 n2 = 35

x1 = 5 x2 = 7

x1 5 x2 7
pˆ 1   pˆ 2  
n1 35 n 2 35
= .14 = .20

For a 98% Confidence Level, /2 = .01 and z.01 = 2.33

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.14)(.86) (.20)(. 80)


35 35
(.14 - .20) ± 2.33 = -.06 ± .21

-.27 < p1 - p2 < .15

10.37 H0: p1 – p2 = 0

Ha: p1 – p2  0
 P a g e | 134

p̂ p̂
 = .10 = .09 = .06 n1 = 780 n2 = 915
1 2

For a two-tailed test, /2 = .05 and z.05 = + 1.645

n1 pˆ 1  n 2 pˆ 2 780(.09)  915(.06)
p 
n1  n 2 780  915
= .0738

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.09  .06)  (0)

Z 
 1 1  1 1 
p  q   (.0738)(. 9262)  
 n1 n   780 915 
 
= 2.35

Since the observed z = 2.35 > z.05 = 1.645, the decision is to reject the null

hypothesis.
 P a g e | 135

p̂ p̂
10.38 n1 = 850 n2 = 910 = .60 2 = .52
1

For a 95% Confidence Level, /2 = .025 and z.025 = + 1.96

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.60)(. 40) (.52)(.48)


850 910
(.60 - .52) + 1.96 = .08 + .046

.034 < p1 – p2 < .126

10.39 H0: 12 = 22  = .01 n1 = 10 s12 = 562

Ha: 12 < 22 n2 = 12 s22 = 1013

dfnum = 12 - 1 = 11 dfdenom = 10 - 1 = 9

Table F.01,10,9 = 5.26

 P a g e | 136
2
s2 1013
2

s1 562
F = = 1.80

Since the observed F = 1.80 < F.01,10,9 = 5.26, the decision is to fail to
reject the null hypothesis.

10.40 H0: 12 = 22  = .05 n1 = 5 s1 = 4.68

Ha: 12  22 n2 = 19 s2 = 2.78

dfnum = 5 - 1 = 4 dfdenom = 19 - 1 = 18

The critical table F values are: F.025,4,18 = 3.61 F.95,18,4 = .277

2
s1 (4.68) 2
2

s2 (2.78) 2
F = = 2.83

Since the observed F = 2.83 < F.025,4,18 = 3.61, the decision is to fail to
reject the null hypothesis.
 P a g e | 137

10.41 City 1 City 2

3.43 3.33

3.40 3.42

3.39 3.39

3.32 3.30

3.39 3.46

3.38 3.39

3.34 3.36

3.38 3.44

3.38 3.37

3.28 3.38

n1 = 10 df1 = 9 n2 = 10 df2 = 9

s12 = .0018989 s22 = .0023378

H0: 12 = 22  = .10 /2 = .05

Ha: 12  22

Upper tail critical F value = F.05,9,9 = 3.18

Lower tail critical F value = F.95,9,9 = 0.314

 P a g e | 138

2
s1 .0018989
2

s2 .0023378
F = = 0.81

Since the observed F = 0.81 is greater than the lower tail critical value
of 0.314 and less than the upper tail critical value of 3.18, the decision
is to fail

to reject the null hypothesis.

 P a g e | 139

10.42 Let Houston = group 1 and Chicago = group 2

1) H0: 12 = 22

Ha: 12  22

2
s1
2
s2
2) F =

3)  = .01

4) df1 = 12 df2 = 10 This is a two-tailed test

The critical table F values are: F.005,12,10 = 5.66 F.995,10,12 = .177

If the observed value is greater than 5.66 or less than .177, the
decision will be to reject the null hypothesis.

5) s12 = 393.4 s22 = 702.7

393.4
702.7
6) F = = 0.56

7) Since F = 0.56 is greater than .177 and less than 5.66,

the decision is to fail to reject the null hypothesis.

8) There is no significant difference in the variances of

number of days between Houston and Chicago.

 P a g e | 140

10.43 H0: 12 = 22  = .05 n1 = 12 s1 = 7.52

Ha: 12 > 22 n2 = 15 s2 = 6.08

dfnum = 12 - 1 = 11 dfdenom = 15 - 1 = 14

The critical table F value is F.05,10,14 = 2.60

2
s1 (7.52) 2
2

s2 (6.08) 2
F= = 1.53

Since the observed F = 1.53 < F.05,10,14 = 2.60, the decision is to fail to
reject the null hypothesis.
 P a g e | 141

10.44 H0: 12 = 22  = .01 n1 = 15 s12 = 91.5

Ha: 12  22 n2 = 15 s22 = 67.3

dfnum = 15 - 1 = 14 dfdenom = 15 - 1 = 14

The critical table F values are: F.005,12,14 = 4.43 F.995,14,12 = .226

2
s1 91.5
2

s2 67.3
F = = 1.36

Since the observed F = 1.36 < F.005,12,14 = 4.43 and > F.995,14,12 = .226,
the decision is to fail to reject the null hypothesis.

10.45 Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

For  = .10 and a two-tailed test, /2 = .05 and z.05 = + 1.645

Sample 1 Sample 2

x x
1 = 138.4 2 = 142.5
 P a g e | 142

1 = 6.71 2 = 8.92

n1 = 48 n2 = 39

( x 1  x 2 )  ( 1   2 ) (138.4  142.5)  (0)


1  2
2 2
(6.71) 2 (8.92)
 
n1 n2 48 39
z = = -2.38

Since the observed value of z = -2.38 is less than the critical value of z
= -1.645, the decision is to reject the null hypothesis. There is a
significant difference in the means of the two populations.
 P a g e | 143

10.46 Sample 1 Sample 2

x x
1 = 34.9 2 = 27.6

12 = 2.97 22 = 3.50

n1 = 34 n2 = 31

For 98% Confidence Level, z.01 = 2.33

2 2
s1 s
( x1  x 2 )  z  2
n1 n2

2.97 3.50

34 31
(34.9 – 27.6) + 2.33 = 7.3 + 1.04

6.26 <  1 -  2 < 8.34

10.47 Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 > 0

Sample 1 Sample 2

x x
= 2.06
1 2 = 1.93
 P a g e | 144

s12 = .176 s22 = .143

n1 = 12 n2 = 15  = .05

This is a one-tailed test with df = 12 + 15 - 2 = 25. The critical value is

t.05,25 = 1.708. If the observed value is greater than 1.708, the decision
will be to reject the null hypothesis.

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t =

(2.06  1.93)  (0)

(.176)(11)  (.143)(14) 1 1

25 12 15
t = = 0.85

Since the observed value of t = 0.85 is less than the critical value of t
= 1.708, the decision is to fail to reject the null hypothesis. The
mean for population one is not significantly greater than the mean for
population two.

10.48 Sample 1 Sample 2

x x
1 = 74.6 2 = 70.9

s12 = 10.5 s22 = 11.4

n1 = 18 n2 = 19
 P a g e | 145

For 95% confidence, /2 = .025.

Using df = 18 + 19 - 2 = 35, t30,.025 = 2.042

2 2
s1 (n1  1)  s 2 (n2  1) 1 1
( x1  x 2 )  t 
n1  n 2  2 n1 n2

(10.5)(17)  (11.4)(18) 1 1

18  19  2 18 19
(74.6 – 70.9) + 2.042

3.7 + 2.22

1.48 <  1 -  2 < 5.92

10.49 Ho: D = 0  = .01

Ha: D < 0

d
n = 21 df = 20 = -1.16 sd = 1.01

The critical t.01,20 = -2.528. If the observed t is less than -2.528, then
the decision will be to reject the null hypothesis.
 P a g e | 146

d  D  1.16  0

sd 1.01
n 21
t = = -5.26

Since the observed value of t = -5.26 is less than the critical t value of
-2.528, the decision is to reject the null hypothesis. The population
difference is less

than zero.
 P a g e | 147

10.50 Respondent Before After d

1 47 63 -16

2 33 35 -2

3 38 36 2

4 50 56 -6

5 39 44 -5

6 27 29 -2

7 35 32 3

8 46 54 -8

9 41 47 -6

d
= -4.44 sd = 5.703 df = 8

For a 99% Confidence Level, /2 = .005 and t8,.005 = 3.355

sd 5.703
d t
n 9
= -4.44 + 3.355 = -4.44 + 6.38

-10.82 < D < 1.94

10.51 Ho: p1 - p2 = 0  = .05 /2 = .025

Ha: p1 - p2  0 z.025 = + 1.96

 P a g e | 148

If the observed value of z is greater than 1.96 or less than -1.96, then
the decision will be to reject the null hypothesis.

Sample 1 Sample 2

x1 = 345 x2 = 421

n1 = 783 n2 = 896

x1  x 2 345  421
p 
n1  n2 783  896
= .4562

x1 345 x 2 421
pˆ 1   pˆ 2  
n1 783 n2 896
= .4406 = .4699

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.4406  .4699)  (0)

z 
 1 1  1 1 
p  q   (.4562)(. 5438)  
 n1 n   783 896 
 
= -1.20

Since the observed value of z = -1.20 is greater than -1.96, the decision
is to fail to reject the null hypothesis. There is no significant difference.

10.52 Sample 1 Sample 2

n1 = 409 n2 = 378

p̂ p̂
1 = .71 2 = .67
 P a g e | 149

For a 99% Confidence Level, /2 = .005 and z.005 = 2.575

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.71)(. 29) (.67)(. 33)


409 378
(.71 - .67) + 2.575 = .04 ± .085

-.045 < p1 - p2 < .125

10.53 H0: 12 = 22  = .05 n1 = 8 s12 = 46

Ha: 12  22 n2 = 10 s22 = 37

dfnum = 8 - 1 = 7 dfdenom = 10 - 1 = 9

The critical F values are: F.025,7,9 = 4.20 F.975,9,7 = .238

If the observed value of F is greater than 4.20 or less than .238, then
the decision will be to reject the null hypothesis.
 P a g e | 150
2
s1 46
2

s2 37
F = = 1.24

Since the observed F = 1.24 is less than F.025,7,9 =4.20 and greater than

F.975,9,7 = .238, the decision is to fail to reject the null hypothesis.

There is no significant difference in the variances of the two
populations.
 P a g e | 151

10.54 Term Whole Life

x x
t = $75,000 w = $45,000

st = $22,000 sw = $15,500

nt = 27 nw = 29

df = 27 + 29 - 2 = 54

For a 95% Confidence Level, /2 = .025 and t.025,50 = 2.009 (used
df=50)

2 2
s1 (n1  1)  s 2 (n2  1) 1 1
( x1  x 2 )  t 
n1  n 2  2 n1 n2

(22,000) 2 (26)  (15,500) 2 (28) 1 1


27  29  2 27 29
(75,000 – 45,000) + 2.009

30,000 ± 10,160.11

19,839.89 < µ1 - µ2 < 40,160.11

10.55 Morning Afternoon d

43 41 2
 P a g e | 152

51 49 2

37 44 -7

24 32 -8

47 46 1

44 42 2

50 47 3

55 51 4

46 49 -3

d
n=9 = -0.444 sd =4.447 df = 9 - 1 = 8

For a 90% Confidence Level: /2 = .05 and t.05,8 = 1.86

sd
d t
n

4.447
9
-0.444 + (1.86) = -0.444 ± 2.757

-3.201 < D < 2.313

10.56 Marketing Accountants

n1 = 400 n2 = 450

x1 = 220 x2 = 216
 P a g e | 153

Ho: p1 - p2 = 0

Ha: p1 - p2 > 0  = .01

The critical table z value is: z.01 = 2.33

220 216
p̂1 400 p̂2 450
= = .55 = = .48

x1  x 2 220  216
p 
n1  n2 400  450
= .513

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.55  .48)  (0)

z 
 1 1  1 1 
p  q   (.513)(.487)  
 n n   400 450 
 1 
= 2.04

Since the observed z = 2.04 is less than z.01 = 2.33, the decision is to
fail to reject

the null hypothesis. There is no significant difference between

marketing

managers and accountants in the proportion who keep track of

obligations “in their head”.

10.57 Accounting Data Entry

 P a g e | 154

n1 = 16 n2 = 14

x x
1 = 26,400 2 = 25,800

s1 = 1,200 s2 = 1,050

H0: 12 = 22

Ha: 12  22  = .05 and 

dfnum = 16 – 1 = 15 dfdenom = 14 – 1 = 13

The critical F values are: F.025,15,13 = 3.05 F.975,15,13 = 0.33

2
s1 1,440,000
2

s2 1,102,500
F = = 1.31

Since the observed F = 1.31 is less than F.025,15,13 = 3.05 and greater
than

F.975,15,13 = 0.33, the decision is to fail to reject the null hypothesis.

10.58 Men Women

n1 = 60 n2 = 41

x x
1 = 631 2 = 848

1 = 100 2 = 100
 P a g e | 155

For a 95% Confidence Level, /2 = .025 and z.025 = 1.96

2 2
s1 s
( x1  x 2 )  z  2
n1 n2

100 2 100 2

60 41
(631 – 848) + 1.96 = -217 ± 39.7

-256.7 < µ1 - µ2 < -177.3

10.59 Ho: µ1 - µ2 = 0  = .01

Ha: µ1 - µ2  0 df = 20 + 24 - 2 = 42

Detroit Charlotte
 P a g e | 156

n1 = 20 n2 = 24

x x
1 = 17.53 2 = 14.89

s1 = 3.2 s2 = 2.7

For two-tail test, /2 = .005 and the critical t.005,40 = ±2.704 (used
df=40)

( x1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n 2  1) 1 1

n1  n2  2 n1 n2
t =

(17.53  14.89)  (0)

(3.2) (19)  (2.7) 2 (23) 1
2
1

42 20 24
t = = 2.97

Since the observed t = 2.97 > t.005,40 = 2.704, the decision is to reject
the null

hypothesis.

10.60 With Fertilizer Without Fertilizer

x x
1 = 38.4 2 = 23.1

1 = 9.8 2 = 7.4
 P a g e | 157

n1 = 35 n2 = 35

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2 > 0

For one-tail test,  = .01 and z.01 = 2.33

( x 1  x 2 )  ( 1   2 ) (38.4  23.1)  (0)


1  2
2 2
(9.8) 2 (7.4) 2
 
n1 n2 35 35
z = = 7.37

Since the observed z = 7.37 > z.01 = 2.33, the decision is to reject the
null

hypothesis.
 P a g e | 158

10.61 Specialty Discount

n1 = 350 n2 = 500

p̂ p̂
1 = .75 2 = .52

For a 90% Confidence Level, /2 = .05 and z.05 = 1.645

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.75)(.25) (.52)(.48)

350 500
(.75 - .52) + 1.645 = .23 ± .053

.177 < p1 - p2 < .283

 P a g e | 159

10.62 H0: 12 = 22  = .01 n1 = 8 n2 = 7

Ha: 12  22 s12 = 72,909 s22 = 129,569

dfnum = 6 dfdenom = 7

The critical F values are: F.005,6,7 = 9.16 F.995,7,6 = .11

2
s1 129,569
2

s2 72,909
F = = 1.78

Since F = 1.78 < F.005,6,7 = 9.16 but also > F.995,7,6 = .11, the decision is
to fail to reject the null hypothesis. There is no difference in the
variances of the shifts.
 P a g e | 160

10.63 Name Brand Store Brand d

54 49 5

55 50 5

59 52 7

53 51 2

54 50 4

61 56 5

51 47 4

53 49 4

d
n=8 = 4.5 sd=1.414 df = 8 - 1 = 7

For a 90% Confidence Level, /2 = .05 and t.05,7 = 1.895

sd
d t
n

1.414
8
4.5 + 1.895 = 4.5 ± .947

3.553 < D < 5.447

 P a g e | 161

10.64 Ho: µ1 - µ2 = 0  = .01

Ha: µ1 - µ2 < 0 df = 23 + 19 - 2 = 40

Wisconsin Tennessee

n1 = 23 n2 = 19

x x
1 = 69.652 2 = 71.7368

s12 = 9.9644 s22 = 4.6491

For one-tail test,  = .01 and the critical t.01,40 = -2.423

( x 1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n2  1) 1 1

n1  n2  2 n1 n2
t =

(69.652  71.7368)  (0)

(9.9644)( 22)  (4.6491)(18) 1 1

40 23 19
t = = -2.44

Since the observed t = -2.44 < t.01,40 = -2.423, the decision is to reject
the null

hypothesis.
 P a g e | 162

10.65 Wednesday Friday d

71 53 18

56 47 9

75 52 23

68 55 13

74 58 16

d
n=5 = 15.8 sd = 5.263 df = 5 - 1 = 4

Ho: D = 0  = .05

Ha: D > 0

For one-tail test,  = .05 and the critical t.05,4 = 2.132

d  D 15.8  0

sd 5.263
n 5
t = = 6.71

Since the observed t = 6.71 > t.05,4 = 2.132, the decision is to reject
the null

hypothesis.

10.66 Ho: p1 - p2 = 0  = .05

Ha: p1 - p2  0
 P a g e | 163

Machine 1 Machine 2

x1 = 38 x2 = 21

n1 = 191 n2 = 202

x1 38 x2 21
pˆ 1   pˆ 2  
n1 191 n2 202
= .199 = .104

n1 pˆ 1  n2 pˆ 2 (.199)(191)  (.104)( 202)

p 
n1  n 2 191  202
= .15

For two-tail, /2 = .025 and the critical z values are: z.025 = ±1.96

( pˆ 1  pˆ 2 )  ( p1  p 2 ) (.199  .104)  (0)

z 
 1 1  1 1 
p  q   (.15)(.85)  
 n1 n   191 202 
 
= 2.64

Since the observed z = 2.64 > zc = 1.96, the decision is to reject the
null

hypothesis.

10.67 Construction Telephone Repair

n1 = 338 n2 = 281

x1 = 297 x2 = 192
 P a g e | 164

x1 297 x 2 192
pˆ 1   pˆ 2  
n1 338 n2 281
= .879 = .683

For a 90% Confidence Level, /2 = .05 and z.05 = 1.645

pˆ 1 qˆ1 pˆ 2 qˆ 2
( pˆ 1  pˆ 2 )  z 
n1 n2

(.879)(. 121) (.683)(.317)


338 281
(.879 - .683) + 1.645 = .196 ± .054

.142 < p1 - p2 < .250

10.68 Aerospace Automobile

n1 = 33 n2 = 35

x x
1 = 12.4 2 = 4.6

1 = 2.9 2 = 1.8

For a 99% Confidence Level, /2 = .005 and z.005 = 2.575

 P a g e | 165

1  2
2 2

( x1  x 2 )  z 
n1 n2

(2.9) 2 (1.8) 2

33 35
(12.4 – 4.6) + 2.575 = 7.8 ± 1.52

6.28 < µ1 - µ2 < 9.32

10.69 Discount Specialty

x x
1 = $47.20 2 = $27.40

1 = $12.45 2 = $9.82

n1 = 60 n2 = 40

Ho: µ1 - µ2 = 0  = .01

Ha: µ1 - µ2  0

For two-tail test, /2 = .005 and zc = ±2.575

( x 1  x 2 )  ( 1   2 ) (47.20  27.40)  (0)


1  2
2 2
(12.45) 2 (9.82) 2
 
n1 n2 60 40
z = = 8.86

Since the observed z = 8.86 > zc = 2.575, the decision is to reject the
null
 P a g e | 166

hypothesis.
 P a g e | 167

10.70 Before After d

12 8 4

7 3 4

10 8 2

16 9 7

8 5 3

d
n=5 = 4.0 sd = 1.8708 df = 5 - 1 =
4

Ho: D = 0  = .01

Ha: D > 0

For one-tail test,  = .01 and the critical t.01,4 = 3.747

d  D 4.0  0

sd 1.8708
n 5
t = = 4.78

Since the observed t = 4.78 > t.01,4 = 3.747, the decision is to reject
the null

hypothesis.

10.71 Ho: µ1 - µ2 = 0  = .01

 P a g e | 168

Ha: µ1 - µ2  0 df = 10 + 6 - 2 = 14

A B___

n1 = 10 n2 = 6

x x
1 = 18.3 2 = 9.667

s12 = 17.122 s22 = 7.467

For two-tail test, /2 = .005 and the critical t.005,14 = ±2.977

( x1  x 2 )  ( 1   2 )
2 2
s1 (n1  1)  s 2 (n 2  1) 1 1

n1  n2  2 n1 n2
t =

(18.3  9.667)  (0)

(17.122)(9)  (7.467)(5) 1 1

14 10 6
t = = 4.52

Since the observed t = 4.52 > t.005,14 = 2.977, the decision is to reject
the null

hypothesis.

10.72 A t test was used to test to determine if Hong Kong has significantly
different

rates than Mumbai. Let group 1 be Hong Kong.

 P a g e | 169

Ho: µ1 - µ2 = 0

Ha: µ1 - µ2  0

x x
n1 = 19 n2 = 23 1 = 130.4 2 = 128.4

s1 = 12.9 s2 = 13.9 98% C.I. and /2 = .01

t = 0.48 with a p-value of .634 which is not significant at of .05. There

is not

enough evidence in these data to declare that there is a difference in

the average

rental rates of the two cities.

10.73 H0: D = 0

Ha: D  0

This is a related measures before and after study. Fourteen people

were involved in the study. Before the treatment, the sample mean
was 4.357 and after the treatment, the mean was 5.214. The higher
number after the treatment indicates that subjects were more likely to
“blow the whistle” after having been through the treatment. The
observed t value was –3.12 which was more extreme than two-tailed
table t value of + 2.16 and as a result, the researcher rejects the null
hypothesis. This is underscored by a p-value of .0081 which is less
than  = .05. The study concludes that there is a significantly higher
likelihood of “blowing the whistle” after the treatment.
 P a g e | 170

10.74 The point estimates from the sample data indicate that in the
northern city the market share is .31078 and in the southern city the
market share is .27013. The point estimate for the difference in the
two proportions of market share are .04065. Since the 99% confidence
interval ranges from -.03936 to +.12067 and zero is in the interval, any
hypothesis testing decision based on this interval would result in failure
to reject the null hypothesis. Alpha is .01 with a two-tailed test. This is
underscored by an observed z value of 1.31 which has an associated p-
value of .191 which, of course, is not significant for any of the usual
values of .
 P a g e | 171

10.75 A test of differences of the variances of the populations of the two

machines is being computed. The hypotheses are:

H0: 12 = 22

Ha: 12 > 22

Twenty-six pipes were measured for sample one and twenty-eight pipes
were measured for sample two. The observed F = 2.0575 is significant
at  = .05 for a one-tailed test since the associated p-value is .034787.
The variance of pipe lengths for machine 1 is significantly greater than
the variance of pipe lengths for machine 2.
 P a g e | 172

Chapter 11
Analysis of Variance and
Design of Experiments

LEARNING OBJECTIVES

The focus of this chapter is learning about the design of experiments and the
analysis of variance thereby enabling you to:

1. Understand the differences between various experiment designs and

when to use them.

2. Compute and interpret the results of a one-way ANOVA.

3. Compute and interpret the results of a random block design.

4. Compute and interpret the results of a two-way ANOVA.

5. Understand and interpret interaction.

6. Know when and how to use multiple comparison techniques.

 P a g e | 173

CHAPTER TEACHING STRATEGY

This important chapter opens the door for students to a broader view
of statistics than they have seen to this time. Through the topic of
experimental designs, the student begins to understand how they can
scientifically set up controlled experiments in which to test certain
hypotheses. They learn about independent and dependent variables. With
the completely randomized design, the student can see how the t test for two
independent samples can be expanded to include three or more samples by
using analysis of variance. This is something that some of the more curious
students were probably wondering about in chapter 10. Through the
randomized block design and the factorial designs, the student can
understand how we can analyze not only multiple categories of one variable,
but we can simultaneously analyze multiple variables with several categories
each. Thus, this chapter affords the instructor an opportunity to help the
student develop a structure for statistical analysis.

In this chapter, we emphasize that the total sum of squares in a given

problem do not change. In the completely randomized design, the total sums
of squares are parceled into between treatments sum of squares and error
sum of squares. By using a blocking design when there is significant
blocking, the blocking effects are removed from the error effect, which
reduces the size of the mean square error and can potentially create a more
powerful test of the treatment. A similar thing happens in the two-way
factorial design when one significant treatment variable siphons off the sum
of squares from the error term that reduces the mean square error and
creates the potential for a more powerful test of the other treatment variable.

In presenting the random block design in this chapter, the emphasis is

on determining if the F value for the treatment variable is significant or not.
There is a de- emphasis on examining the F value of the blocking effects.
However, if the blocking effects are not significant, the random block design
may be a less powerful analysis of the treatment effects. If the blocking
effects are not significant, even though the error sum of squares is reduced,
the mean square error might increase because the blocking effects may
reduce the degrees of freedom error in a proportionaly greater amount. This
 P a g e | 174

might result in a smaller treatment F value than would occur in a completely

randomized design. The repeated-measures design is shown in the chapter
as a special case of the random block design.

In factorial designs, if there are multiple values in the cells, it is

possible to analyze interaction effects. Random block designs do not have
multiple values in cells and therefore interaction effects cannot be calculated.
It is emphasized in this chapter that if significant interaction occurs, then the
main effects analysis are confounded and should not be analyzed in the usual
manner. There are various philosophies about how to handle significant
interaction but are beyond the scope of this chapter. The main factorial
example problem in the chapter was created to have no significant
interaction so that the student can learn how to analyze main effects. The
demonstration problem has significant interaction and these interactions are
displayed graphically for the student to see. You might consider taking this
same problem and graphing the interactions using row effects along the x
axis and graphing the column means for the student to see.

There are a number of multiple comparison tests available. In this

text, one of the more well-known tests, Tukey's HSD, is featured in the case of
equal sample sizes. When sample sizes are unequal, a variation on Tukey’s
HSD, the Tukey-Kramer test, is used. MINITAB uses the Tukey test as one of
its options under multiple comparisons and uses the Tukey-Kramer test for
unequal sample sizes. Tukey's HSD is one of the more powerful multiple
comparison tests but protects less against Type I errors than some of the
other tests.
 P a g e | 175

CHAPTER OUTLINE

11.1 Introduction to Design of Experiments

11.2 The Completely Randomized Design (One-Way ANOVA)

One-Way Analysis of Variance

Reading the F Distribution Table

Using the Computer for One-Way ANOVA

Comparison of F and t Values

11.3 Multiple Comparison Tests

Tukey's Honestly Significant Difference (HSD) Test: The Case of Equal

Sample

Sizes

Using the Computer to Do Multiple Comparisons

Tukey-Kramer Procedure: The Case of Unequal Sample Sizes

11.4 The Randomized Block Design

Using the Computer to Analyze Randomized Block Designs

11.5 A Factorial Design (Two-Way ANOVA)

Advantages of the Factorial Design

Factorial Designs with Two Treatments

Applications

Statistically Testing the Factorial Design

Interaction
 P a g e | 176

Using a Computer to Do a Two-Way ANOVA

KEY TERMS

a posteriori Factors

a priori Independent Variable

Analysis of Variance (ANOVA) Interaction

Blocking Variable Levels

Classification Variables Multiple Comparisons

Classifications One-way Analysis of

Variance

Completely Randomized Design Post-hoc

Concomitant Variables Randomized Block Design

Confounding Variables Repeated Measures Design

Dependent Variable Treatment Variable

Experimental Design Tukey-Kramer Procedure

F Distribution Tukey’s HSD Test

F Value Two-way Analysis of Variance

Factorial Design

SOLUTIONS TO PROBLEMS IN CHAPTER 11

 P a g e | 177

11.1 a) Time Period, Market Condition, Day of the Week, Season of the Year

b) Time Period - 4 P.M. to 5 P.M. and 5 P.M. to 6 P.M.

Market Condition - Bull Market and Bear Market

Day of the Week - Monday, Tuesday, Wednesday, Thursday, Friday

Season of the Year – Summer, Winter, Fall, Spring

c) Volume, Value of the Dow Jones Average, Earnings of Investment

Houses

11.2 a) Type of 737, Age of the plane, Number of Landings per Week of the
plane,

City that the plane is based

b) Type of 737 - Type I, Type II, Type III

Age of plane - 0-2 y, 3-5 y, 6-10 y, over 10 y

Number of Flights per Week - 0-5, 6-10, over 10

City - Dallas, Houston, Phoenix, Detroit

c) Average annual maintenance costs, Number of annual hours spent

on

maintenance
 P a g e | 178

11.3 a) Type of Card, Age of User, Economic Class of Cardholder, Geographic

Region

b) Type of Card - Mastercard, Visa, Discover, American Express

Age of User - 21-25 y, 26-32 y, 33-40 y, 41-50 y, over 50

Economic Class - Lower, Middle, Upper

Geographic Region - NE, South, MW, West

c) Average number of card usages per person per month,

Average balance due on the card, Average per expenditure

per person,

Number of cards possessed per person

11.4 Average dollar expenditure per day/night, Age of adult registering the
family, Number of days stay (consecutive)
 P a g e | 179

11.5 Source df SS MS F__

Treatment 2 22.20 11.10 11.07

Error 14 14.03 1.00______

Total 16 36.24

 = .05 Critical F.05,2,14 = 3.74

Since the observed F = 11.07 > F.05,2,14 = 3.74, the decision is to reject
the null hypothesis.

11.6 Source df SS MS F__

Treatment 4 93.77 23.44 15.82

Error 18 26.67 1.48______

Total 22 120.43

 = .01 Critical F.01,4,18 = 4.58

Since the observed F = 15.82 > F.01,4,18 = 4.58, the decision is to reject
the null

hypothesis.
 P a g e | 180

11.7 Source df SS MS F_

Treatment 3 544.2 181.4 13.00

Error 12 167.5 14.0______

Total 15 711.8

 = .01 Critical F.01,3,12 = 5.95

Since the observed F = 13.00 > F.01,3,12 = 5.95, the decision is to reject
the null hypothesis.
 P a g e | 181

11.8 Source df SS MS F__

Treatment 1 64.29 64.29 17.76

Error 12 43.43 3.62______

Total 13 107.71

 = .05 Critical F.05,1,12 = 4.75

Since the observed F = 17.76 > F.05,1,12 = 4.75, the decision is to reject
the null hypothesis.

Observed t value using t test:

1 2

n1 = 7 n2 = 7

x x
1 = 29 2 = 24.71429

s12 = 3 s22 = 4.238095

(29  24.71429)  (0)

3(6)  (4.238095)( 6) 1 1

772 7 7
t = = 4.214

F  17.76
Also, t = = 4.214
 P a g e | 182

11.9 Source SS df MS F__

Treatment 583.39 4 145.8475 7.50

Error 972.18 50 19.4436______

Total 1,555.57 54

11.10 Source SS df MS F _

Treatment 29.64 2 14.820 3.03

Error 68.42 14 4.887___ __

Total 98.06 16

F.05,2,14 = 3.74

Since the observed F = 3.03 < F.05,2,14 = 3.74, the decision is to fail to
reject

the null hypothesis

11.11 Source df SS MS F__

 P a g e | 183

Treatment 3 .007076 .002359 10.10

Error 15 .003503 .000234________

Total 18 .010579

 = .01 Critical F.01,3,15 = 5.42

Since the observed F = 10.10 > F.01,3,15 = 5.42, the decision is to reject
the null

hypothesis.

11.12 Source df SS MS F__

Treatment 2 180700000 90350000 92.67

Error 12 11699999 975000_________

Total 14 192400000

 = .01 Critical F.01,2,12 = 6.93

Since the observed F = 92.67 > F.01,2,12 = 6.93, the decision is to reject
the null

hypothesis.
 P a g e | 184

11.13 Source df SS MS F___

Treatment 2 29.61 14.80 11.76

Error 15 18.89 1.26________

Total 17 48.50

 = .05 Critical F.05,2,15 = 3.68

Since the observed F = 11.76 > F.05,2,15 = 3.68, the decison is to reject
the null

hypothesis.
 P a g e | 185

11.14 Source df SS MS F__

Treatment 3 456630 152210 11.03

Error 16 220770 13798_______

Total 19 677400

 = .05 Critical F.05,3,16 = 3.24

Since the observed F = 11.03 > F.05,3,16 = 3.24, the decision is to reject
the null

hypothesis.

11.15 There are 4 treatment levels. The sample sizes are 18, 15, 21, and
11. The F

value is 2.95 with a p-value of .04. There is an overall significant

difference at

alpha of .05. The means are 226.73, 238.79, 232.58, and 239.82.

11.16 The independent variable for this study was plant with five
classification levels
(the five plants). There were a total of 43 workers who participated in
the study. The dependent variable was number of hours worked per
 P a g e | 186

week. An observed F value of 3.10 was obtained with an associated p-

value of .026595. With an alpha of .05, there was a significant
overall difference in the average number of hours worked per week
by plant. A cursory glance at the plant averages revealed that workers
at plant 3 averaged 61.47 hours per week (highest number) while
workers at plant 4 averaged 49.20 (lowest number).

11.17C = 6 MSE = .3352  = .05 N = 46

x x
q.05,6,40 = 4.23 n3 = 8 n6 = 7 3 = 15.85 6 = 17.2

.3352  1 1 
  
2  8 7
HSD = 4.23 = 0.896

x 3  x 6  15.85  17.21
= 1.36

Since 1.36 > 0.896, there is a significant difference between the

means of

groups 3 and 6.

11.18 C = 4 n=6 N = 24 dferror = N - C = 24 - 4 = 20  = .05

 P a g e | 187

MSE = 2.389 q.05,4,20 = 3.96

MSE 2.389
n 6
HSD = q = (3.96) = 2.50

11.19 C = 3 MSE = 1.002381  = .05 N = 17 N - C = 14

x x
q.05,3,14 = 3.70 n1 = 6 n2 = 5 1 =2 2 = 4.6

1.002381  1 1 
  
2  6 5
HSD = 3.70 = 1.586

x 1  x 2  2  4 .6
= 2.6

Since 2.6 > 1.586, there is a significant difference between the

means of

groups 1 and 2.
 P a g e | 188

11.20 From problem 11.6, MSE = 1.481481 C=5 N = 23 N – C

= 18

n2 = 5 n4 = 5  = .01 q.01,5,18 = 5.38

1.481481  1 1 
  
2  5 5
HSD = 5.38 = 2.93

x x
2 = 10 4 = 16

x 3  x 6  10  16
= 6

Since 6 > 2.93, there is a significant difference in the means of

groups 2 and 4.
 P a g e | 189

11.21 N = 16 n=4 C=4 N - C = 12 MSE = 13.95833

q.01,4,12 = 5.50

MSE 13.95833
n 4
HSD = q = 5.50 = 10.27

x x x x
1 = 115.25 2 = 125.25 3 = 131.5 4 = 122.5

x x
1 and 3 are the only pair that are significantly different
using the

HSD test.
 P a g e | 190

11.22 n = 7 C=2 MSE = 3.619048 N = 14 N - C = 14 - 2 =

12

 = .05 q.05,2,12 = 3.08

MSE 3.619048
n 7
HSD = q = 3.08 = 2.215

x x
1 = 29 and 2 = 24.71429

x x
Since 1 - 2 = 4.28571 > HSD = 2.215, the decision is to reject the
null

hypothesis.
 P a g e | 191

11.23 C = 4 MSE = .000234  = .01 N = 19 N – C = 15

q.01,4,15 = 5.25 n1 = 4 n2 = 6 n3 = 5 n4 = 4

x x x x
1 = 4.03, 2 = 4.001667, 3 = 3.974, 4 = 4.005

.000234  1 1 
  
2  4 6
HSD1,2 = 5.25 = .0367

.000234  1 1 
  
2  4 5
HSD1,3 = 5.25 = .0381

.000234  1 1 
  
2  4 4
HSD1,4 = 5.25 = .0402

.000234  1 1 
  
2  6 5
HSD2,3 = 5.25 = .0344

.000234  1 1 
  
2  6 4
HSD2,4 = 5.25 = .0367
 P a g e | 192

.000234  1 1 
  
2  5 4
HSD3,4 = 5.25 = .0381

x1  x 3
= .056

This is the only pair of means that are significantly different.

 P a g e | 193

11.24  = .01 C=3 n=5 N = 15 N – C = 12

MSE = 975,000 q.01,3,12 = 5.04

MSE 975,000
n 5
HSD = q = 5.04 = 2,225.6

x x x
1 = 40,900 2 = 49,400 3 = 45,300

x1  x 2
= 8,500

x1  x 3
= 4,400

x 2  x3
= 4,100

Using Tukey's HSD, all three pairwise comparisons are

significantly different.

11.25  = .05 C=3 N = 18 N - C = 15 MSE = 1.259365

q.05,3,15 = 3.67 n1 = 5 n2 = 7 n3 = 6
 P a g e | 194

x x x
1 = 7.6 2 = 8.8571 3 = 5.8333

1.259365  1 1 
  
2  5 7
HSD1,2 = 3.67 = 1.705

1.259365  1 1 
  
2  5 6
HSD1,3 = 3.67 = 1.763

1.259365  1 1 
  
2  7 6
HSD2,3 = 3.67 = 1.620

x1  x 3
= 1.767 (is significant)

x 2  x3
= 3.024 (is significant)
 P a g e | 195

11.26  = .05 n=5 C=4 N = 20 N - C = 16 MSE =

13,798.13

x x x x
1 = 591 2 = 350 3 = 776 4 = 563

MSE 13,798.13
n 5
HSD = q = 4.05 = 212.76

x1  x 2 x1  x 3 x1  x 4
= 241 = 185 = 28

x2  x3 x2  x4 x3  x4
= 426 = 213 = 213

Using Tukey's HSD = 212.76, means 1 and 2, means 2 and 3,

means 2 and 4,

and means 3 and 4 are significantly different.

11.27  = .05. There were five plants and ten pairwise comparisons. The
MINITAB

output reveals that the only significant pairwise difference is

between plant 2 and
 P a g e | 196

plant 3 where the reported confidence interval (0.180 to 22.460)

contains the same
sign throughout indicating that 0 is not in the interval. Since zero is
not in the
interval, then we are 95% confident that there is a pairwise difference
significantly different from zero. The lower and upper values for all
other
confidence intervals have different signs indicating that zero is
included in the
interval. This indicates that the difference in the means for these pairs
might be
zero.
 P a g e | 197

11.28 H0: µ1 = µ2 = µ3 = µ4

Ha: At least one treatment mean is different from the others

Source df SS MS F__

Treatment 3 62.95 20.9833 5.56

Blocks 4 257.50 64.3750 17.05

Error 12 45.30 3.7750______

Total 19 365.75

 = .05 Critical F.05,3,12 = 3.49 for treatments

For treatments, the observed F = 5.56 > F.05,3,12 = 3.49, the decision is
to

reject the null hypothesis.

11.29 H0: µ1 = µ2 = µ3

Ha: At least one treatment mean is different from the others

Source df SS MS F_

Treatment 2 .001717 .000858 1.48

Blocks 3 .076867 .025622 44.13

Error 6 .003483 .000581_______

Total 11 .082067
 P a g e | 198

 = .01 Critical F.01,2,6 = 10.92 for treatments

For treatments, the observed F = 1.48 < F.01,2,6 = 10.92 and the
decision is to

fail to reject the null hypothesis.

11.30 Source df SS MS F__

Treatment 5 2477.53 495.506 1.91

Blocks 9 3180.48 353.387 1.36

Error 45 11661.38 259.142______

Total 59 17319.39

 = .05 Critical F.05,5,45 = 2.45 for treatments

For treatments, the observed F = 1.91 < F.05,5,45 = 2.45 and decision is
to fail to

reject the null hypothesis.

11.31 Source df SS MS F__

Treatment 3 199.48 66.493 3.90

Blocks 6 265.24 44.207 2.60

Error 18 306.59 17.033______

 P a g e | 199

Total 27 771.31

 = .01 Critical F.01,3,18 = 5.09 for treatments

For treatments, the observed F = 3.90 < F.01,3,18 = 5.09 and the decision
is to

fail to reject the null hypothesis.

11.32 Source df SS MS F__

Treatment 3 2302.5 767.5000 15.67

Blocks 9 5402.5 600.2778 12.26

Error 27 1322.5 48.9815____ __

Total 39 9027.5

 = .05 Critical F.05,3,27 = 2.96 for treatments

For treatments, the observed F = 15.67 > F.05,3,27 = 2.96 and the
decision is to

reject the null hypothesis.

 P a g e | 200

11.33 Source df SS MS F__

Treatment 2 64.5333 32.2667 15.37

Blocks 4 137.6000 34.4000 16.38

Error 8 16.8000 2.1000_ _____

Total 14 218.9300

 = .01 Critical F.01,2,8 = 8.65 for treatments

For treatments, the observed F = 15.37 > F.01,2,8 = 8.65 and the
decision is to

reject the null hypothesis.

 P a g e | 201

11.34 This is a randomized block design with 3 treatments (machines) and 5

block levels (operators). The F for treatments is 6.72 with a p-value
of .019. There is a significant difference in machines at  = .05. The
F for blocking effects is 0.22 with a p-value of .807. There are no
significant blocking effects. The blocking effects reduced the power of
the treatment effects since the blocking effects were not significant.

11.35 The p value for Phone Type, .00018, indicates that there is an overall significant
difference in treatment means at alpha .001. The lengths of calls differ according to type
of telephone used. The p-value for managers, .00028, indicates that there is an overall
difference in block means at alpha .001. The lengths of calls differ according to Manager.
The significant blocking effects have improved the power of the F test for treatments.

11.36 This is a two-way factorial design with two independent variables and
one dependent variable. It is 2x4 in that there are two row treatment
levels and four column treatment levels. Since there are three
measurements per cell, interaction can be analyzed.

dfrow treatment = 1 dfcolumn treatment = 3 dfinteraction = 3 dferror = 16

dftotal = 23

11.37 This is a two-way factorial design with two independent variables and
one dependent variable. It is 4x3 in that there are four treatment
 P a g e | 202

levels and three column treatment levels. Since there are two
measurements per cell, interaction can be analyzed.

dfrow treatment = 3 dfcolumn treatment = 2 dfinteraction = 6 dferror = 12 dftotal

= 23

11.38 Source df SS MS F__

Row 3 126.98 42.327 3.46

Column 4 37.49 9.373 0.77

Interaction 12 380.82 31.735 2.60

Error 60 733.65 12.228______

Total 79 1278.94

 = .05

Critical F.05,3,60 = 2.76 for rows. For rows, the observed F = 3.46 >
F.05,3,60 = 2.76

and the decision is to reject the null hypothesis.

Critical F.05,4,60 = 2.53 for columns. For columns, the observed F = 0.77
<

F.05,4,60 = 2.53 and the decision is to fail to reject the null

hypothesis.

Critical F.05,12,60 = 1.92 for interaction. For interaction, the observed F =

2.60 >
 P a g e | 203

F.05,12,60 = 1.92 and the decision is to reject the null hypothesis.

Since there is significant interaction, the researcher should exercise

extreme

caution in analyzing the "significant" row effects.

11.39 Source df SS MS F__

Row 1 1.047 1.047 2.40

Column 3 3.844 1.281 2.94

Interaction 3 0.773 0.258 0.59

Error 16 6.968 0.436______

Total 23 12.632

 = .05

Critical F.05,1,16 = 4.49 for rows. For rows, the observed F = 2.40 <
F.05,1,16 = 4.49

and decision is to fail to reject the null hypothesis.

Critical F.05,3,16 = 3.24 for columns. For columns, the observed F = 2.94
<

F.05,3,16 = 3.24 and the decision is to fail to reject the null

hypothesis.
 P a g e | 204

Critical F.05,3,16 = 3.24 for interaction. For interaction, the observed F =

0.59 <

F.05,3,16 = 3.24 and the decision is to fail to reject the null

hypothesis.
 P a g e | 205

11.40 Source df SS MS F___

Row 1 60.750 60.750 38.37

Column 2 14.000 7.000 4.42

Interaction 2 2.000 1.000 0.63

Error 6 9.500 1.583________

Total 11 86.250

 = .01

Critical F.01,1,6 = 13.75 for rows. For rows, the observed F = 38.37 >

F.01,1,6 = 13.75 and the decision is to reject the null hypothesis.

Critical F.01,2,6 = 10.92 for columns. For columns, the observed F = 4.42
<

F.01,2,6 = 10.92 and the decision is to fail to reject the null

hypothesis.

Critical F.01,2,6 = 10.92 for interaction. For interaction, the observed F =

0.63 <

F.01,2,6 = 10.92 and the decision is to fail to reject the null

hypothesis.

11.41 Source df SS MS F__

Treatment 1 1 1.24031 1.24031 63.67

Treatment 2 3 5.09844 1.69948 87.25
Interaction 3 0.12094 0.04031 2.07
 P a g e | 206

Error 24 0.46750 0.01948______

Total 31 6.92719

 = .05

Critical F.05,1,24 = 4.26 for treatment 1. For treatment 1, the observed F

= 63.67 >

F.05,1,24 = 4.26 and the decision is to reject the null hypothesis.

Critical F.05,3,24 = 3.01 for treatment 2. For treatment 2, the observed F

= 87.25 >

F.05,3,24 = 3.01 and the decision is to reject the null hypothesis.

Critical F.05,3,24 = 3.01 for interaction. For interaction, the observed F =

2.07 <

F.05,3,24 = 3.01 and the decision is to fail to reject the null

hypothesis.
 P a g e | 207

11.42 Source df SS MS F__

Age 3 42.4583 14.1528 14.77

No. Children 2 49.0833 24.5417 25.61

Interaction 6 4.9167 0.8194 0.86

Error 12 11.5000 0.9583______

Total 23 107.9583

 = .05

Critical F.05,3,12 = 3.49 for Age. For Age, the observed F = 14.77 >

F.05,3,12 = 3.49 and the decision is to reject the null hypothesis.

Critical F.05,2,12 = 3.89 for No. Children. For No. Children, the observed

F = 25.61 > F.05,2,12 = 3.89 and the decision is to reject the null
hypothesis.

Critical F.05,6,12 = 3.00 for interaction. For interaction, the observed F =

0.86 <

F.05,6,12 = 3.00 and fail to reject the null hypothesis.

11.43 Source df SS MS F__

Location 2 1736.22 868.11 34.31
Competitors 3 1078.33 359.44 14.20
Interaction 6 503.33 83.89 3.32
Error 24 607.33 25.31_______

Total 35 3925.22
 P a g e | 208

 = .05

Critical F.05,2,24 = 3.40 for rows. For rows, the observed F = 34.31 >

F.05,2,24 = 3.40 and the decision is to reject the null hypothesis.

Critical F.05,3,24 = 3.01 for columns. For columns, the observed F =

14.20 >

F.05,3,24 = 3.01 and decision is to reject the null hypothesis.

Critical F.05,6,24 = 2.51 for interaction. For interaction, the observed F =

3.32 >

F.05,6,24 = 2.51 and the decision is to reject the null hypothesis.

Note: There is significant interaction in this study. This may confound

the

interpretation of the main effects, Location and Number of

Competitors.
 P a g e | 209

11.44 This two-way design has 3 row treatments and 5 column treatments.
There are 45 total observations with 3 in each cell.

MS R 46.16

MS E 3.49
FR = = 13.23

p-value = .000 and the decision is to reject the null hypothesis for
rows.

MSC 249.70

MS E 3.49
FC = = 71.57

p-value = .000 and the decision is to reject the null hypothesis for
columns.

MS I 55.27

MS E 3.49
FI = = 15.84

p-value = .000 and the decision is to reject the null hypothesis for
interaction.

Because there is significant interaction, the analysis of main effects

is confounded. The graph of means displays the crossing patterns of
the line segments indicating the presence of interaction.
 P a g e | 210

11.45 The null hypotheses are that there are no interaction effects, that there
are no significant differences in the means of the valve openings by
machine, and that there are no significant differences in the means of
the valve openings by shift. Since the p-value for interaction effects
is .876, there are no significant interaction effects and that is good
since significant interaction effects would confound that study. The p-
value for columns (shifts) is .008 indicating that column effects are
significant at alpha of .01. There is a significant difference in the mean
valve opening according to shift. No multiple comparisons are given in
the output. However, an examination of the shift means indicates that
the mean valve opening on shift one was the largest at 6.47 followed
by shift three with 6.3 and shift two with 6.25. The p-value for rows
(machines) is .937 and that is not significant.
 P a g e | 211

11.46 This two-way factorial design has 3 rows and 3 columns with three
observations
per cell. The observed F value for rows is 0.19, for columns is 1.19,
and for

interaction is 1.40. Using an alpha of .05, the critical F value for rows
and columns (same df) is F2,18,.05 = 3.55. Neither the observed F value
for rows nor the observed F value for columns is significant. The
critical F value for interaction is F4,18,.05 = 2.93. There is no significant
interaction.

11.47 Source df SS MS F__

Treatment 3 66.69 22.23 8.82

Error 12 30.25 2.52______

Total 15 96.94

 = .05 Critical F.05,3,12 = 3.49

Since the treatment F = 8.82 > F.05,3,12 = 3.49, the decision is to reject
the null

hypothesis.
 P a g e | 212

For Tukey's HSD:

MSE = 2.52 n=4 N = 16 C=4 N - C = 12

q.05,4,12 = 4.20

MSE 2.52
n 4
HSD = q = (4.20) = 3.33

x x x x
1 = 12 2 = 7.75 3 = 13.25 4 = 11.25

Using HSD of 3.33, there are significant pairwise differences between

means 1 and 2, means 2 and 3, and means 2 and 4.

11.48 Source df SS MS F__

Treatment 6 68.19 11.365 0.87

Error 19 249.61 13.137______

Total 25 317.80
 P a g e | 213

11.49 Source df SS MS F__

Treatment 5 210 42.000 2.31

Error 36 655 18.194______

Total 41 865

11.50 Source df SS MS F__

Treatment 2 150.91 75.46 16.19

Error 22 102.53 4.66________

Total 24 253.44

 = .01 Critical F.01,2,22 = 5.72

Since the observed F = 16.19 > F.01,2,22 = 5.72, the decision is to reject
the null

hypothesis.

x x x
1 = 9.200 2 = 14.250 3 = 8.714286

n1 = 10 n2 = 8 n3 = 7
 P a g e | 214

MSE = 4.66 C=3 N = 25 N - C = 22

 = .01 q.01,3,22 = 4.64

4.66  1 1 
  
2  10 8 
HSD1,2 = 4.64 = 3.36

4.66  1 1 
  
2  10 7 
HSD1,3 = 4.64 = 3.49

4.66  1 1 
  
2  8 7
HSD2,3 = 4.64 = 3.67

x1  x 2 x 2  x3
= 5.05 and = 5.5357 are significantly
different at  = .01
 P a g e | 215

11.51 This design is a repeated-measures type random block design. There

is one

treatment variable with three levels. There is one blocking variable

with six

people in it (six levels). The degrees of freedom treatment are two.

The degrees

of freedom block are five. The error degrees of freedom are ten. The
total

degrees of freedom are seventeen. There is one dependent variable.

11.52 Source df SS MS F__

Treatment 3 20,994 6998.00 5.58

Blocks 9 16,453 1828.11 1.46

Error 27 33,891 1255.22_____

Total 39 71,338

 = .05 Critical F.05,3,27 = 2.96 for treatments

Since the observed F = 5.58 > F.05,3,27 = 2.96 for treatments, the
decision is to

reject the null hypothesis.

 P a g e | 216

11.53 Source df SS MS F__

Treatment 3 240.125 80.042 31.51

Blocks 5 548.708 109.742 43.20

Error 15 38.125 2.542_ _____

Total 23

 = .05 Critical F.05,3,15 = 3.29 for treatments

Since for treatments the observed F = 31.51 > F.05,3,15 = 3.29, the
decision is to

reject the null hypothesis.

For Tukey's HSD:

Ignoring the blocking effects, the sum of squares blocking and sum of
squares error are combined together for a new SS error = 548.708 +
38.125 = 586.833. Combining the degrees of freedom error and
blocking yields a new dferror = 20. Using these new figures, we compute
a new mean square error, MSE = (586.833/20) = 29.34165.

n=6 C=4 N = 24 N - C = 20 q.05,4,20 = 3.96

 P a g e | 217

MSE 29.34165
n 6
HSD = q = (3.96) = 8.757

x x x x
1 = 16.667 2 = 12.333 3 = 12.333 4 = 19.833

None of the pairs of means are significantly different using Tukey's HSD
= 8.757.

This may be due in part to the fact that we compared means by folding
the

blocking effects back into error and the blocking effects were highly

significant.

11.54 Source df SS MS F__

Treatment 1 4 29.13 7.2825 1.98

Treatment 2 1 12.67 12.6700 3.44

Interaction 4 73.49 18.3725 4.99

Error 30 110.38 3.6793______

Total 39 225.67

 = .05
 P a g e | 218

Critical F.05,4,30 = 2.69 for treatment 1. For treatment 1, the observed F

= 1.98 <

F.05,4,30 = 2.69 and the decision is to fail to reject the null

hypothesis.

Critical F.05,1,30 = 4.17 for treatment 2. For treatment 2 observed F =

3.44 <

F.05,1,30 = 4.17 and the decision is to fail to reject the null

hypothesis.

Critical F.05,4,30 = 2.69 for interaction. For interaction, the observed F =

4.99 >

F.05,4,30 = 2.69 and the decision is to reject the null hypothesis.

Since there are significant interaction effects, examination of the main

effects

should not be done in the usual manner. However, in this case, there
are no

significant treatment effects anyway.

 P a g e | 219

11.55 Source df SS MS F___

Treatment 2 3 257.889 85.963 38.21

Treatment 1 2 1.056 0.528 0.23

Interaction 6 17.611 2.935 1.30

Error 24 54.000 2.250________

Total 35 330.556

 = .01

Critical F.01,3,24 = 4.72 for treatment 2. For the treatment 2 effects, the
observed

F = 38.21 > F.01,3,24 = 4.72 and the decision is to reject the null
hypothesis.

Critical F.01,2,24 = 5.61 for Treatment 1. For the treatment 1 effects, the
observed

F = 0.23 < F.01,2,24 = 5.61 and the decision is to fail to reject the null
hypothesis.

Critical F.01,6,24 = 3.67 for interaction. For the interaction effects, the
observed

F = 1.30 < F.01,6,24 = 3.67 and the decision is to fail to reject the null
hypothesis.
 P a g e | 220

11.56 Source df SS MS F___

Age 2 49.3889 24.6944 38.65

Column 3 1.2222 0.4074 0.64

Interaction 6 1.2778 0.2130 0.33

Error 24 15.3333 0.6389_______

Total 35 67.2222

 = .05

Critical F.05,2,24 = 3.40 for Age. For the age effects, the observed F =
38.65 >

F.05,2,24 = 3.40 and the decision is to reject the null hypothesis.

Critical F.05,3,24 = 3.01 for Region. For the region effects, the observed F
= 0.64

< F.05,3,24 = 3.01 and the decision is to fail to reject the null
hypothesis.

Critical F.05,6,24 = 2.51 for interaction. For interaction effects, the

observed

F = 0.33 < F.05,6,24 = 2.51 and the decision is to fail to reject the null
hypothesis.

There are no significant interaction effects. Only the Age effects are
significant.
 P a g e | 221

Computing Tukey's HSD for Age:

x x x
1 = 2.667 2 = 4.917 3 = 2.250

n = 12 C=3 N = 36 N - C = 33

MSE is recomputed by folding together the interaction and column sum

of

squares and degrees of freedom with previous error terms:

MSE = (1.2222 + 1.2778 + 15.3333)/(3 + 6 + 24) = 0.5404

q.05,3,33 = 3.49

MSE .5404
n 12
HSD = q = (3.49) = 0.7406

Using HSD, there are significant pairwise differences between means 1

and 2 and

between means 2 and 3.

Shown below is a graph of the interaction using the cell means by Age.
P a g e | 222
 P a g e | 223

11.57 Source df SS MS F__

Treatment 3 90477679 30159226 7.38

Error 20 81761905 4088095_______

Total 23 172000000

 = .05 Critical F.05,3,20 = 3.10

The treatment F = 7.38 > F.05,3,20 = 3.10 and the decision is to reject
the null

hypothesis.

11.58 Source df SS MS F__

Treatment 2 460,353 230,176 103.70

Blocks 5 33,524 6,705 3.02

Error 10 22,197 2,220_______

Total 17 516,074

 = .01 Critical F.01,2,10 = 7.56 for treatments

Since the treatment observed F = 103.70 > F.01,2,10 = 7.56, the

decision is to

reject the null hypothesis.

 P a g e | 224

11.59 Source df SS MS F__

Treatment 2 9.555 4.777 0.46

Error 18 185.1337 10.285_______

Total 20 194.6885

 = .05 Critical F.05,2,18 = 3.55

Since the treatment F = 0.46 > F.05,2,18 = 3.55, the decision is to fail to
reject the

null hypothesis.

Since there are no significant treatment effects, it would make no

sense to

compute Tukey-Kramer values and do pairwise comparisons.

 P a g e | 225

11.60 Source df SS MS F___

Years 2 4.875 2.437 5.16

Size 3 17.083 5.694 12.06

Interaction 6 2.292 0.382 0.81

Error 36 17.000 0.472_______

Total 47 41.250

 = .05

Critical F.05,2,36 = 3.32 for Years. For Years, the observed F = 5.16 >

F.05,2,36 = 3.32 and the decision is to reject the null hypothesis.

Critical F.05,3,36 = 2.92 for Size. For Size, the observed F = 12.06 >
F.05,3,36 = 2.92

and the decision is to reject the null hypothesis.

Critical F.05,6,36 = 2.42 for interaction. For interaction, the observed F =

0.81 <

F.05,6,36 = 2.42 and the decision is to fail to reject the null

hypothesis.

There are no significant interaction effects. There are significant row

and column

effects at  = .05.
 P a g e | 226

11.61 Source df SS MS F___

Treatment 4 53.400 13.350 13.64

Blocks 7 17.100 2.443 2.50

Error 28 27.400 0.979________

Total 39 97.900

 = .05 Critical F.05,4,28 = 2.71 for treatments

For treatments, the observed F = 13.64 > F.05,4,28 = 2.71 and the
decision is to

reject the null hypothesis.

 P a g e | 227

11.62 This is a one-way ANOVA with four treatment levels. There are 36
observations

in the study. The p-value of .045 indicates that there is a significant

overall

difference in the means at  = .05. An examination of the mean

analysis shows

that the sample sizes are different with sizes of 8, 7, 11, and 10,
respectively. No

multiple comparison technique was used here to conduct pairwise

comparisons.

However, a study of sample means shows that the two most extreme
means are

from levels one and four. These two means would be the most likely
candidates

for multiple comparison tests. Note that the confidence intervals for
means one

and four (shown in the graphical output) are seemingly non-

overlapping

indicating a potentially significant difference.

11.63 Excel reports that this is a two-factor design without replication

indicating that

this is a random block design. Neither the row nor the column p-values
are less

than .05 indicating that there are no significant treatment or blocking

effects in this study. Also displayed in the output to underscore this
conclusion are the observed and critical F values for both treatments
and blocking. In both
 P a g e | 228

cases, the observed value is less than the critical value.

11.64 This is a two-way ANOVA with 5 rows and 2 columns. There are 2
observations
per cell. For rows, FR = 0.98 with a p-value of .461 which is not
significant. For columns, FC = 2.67 with a p-value of .134 which is not
significant. For interaction, FI = 4.65 with a p-value of .022 which is
significant at  = .05. Thus, there are significant interaction effects
and the row and column effects are confounded. An examination of
the interaction plot reveals that most of the lines cross verifying the
finding of significant interaction.

11.65 This is a two-way ANOVA with 4 rows and 3 columns. There are 3
observations

per cell. FR = 4.30 with a p-value of .014 is significant at  = .05. The

null

hypothesis is rejected for rows. FC = 0.53 with a p-value of .594 is not

significant. We fail to reject the null hypothesis for columns. FI = 0.99

with a

p-value of .453 for interaction is not significant. We fail to reject the

null

hypothesis for interaction effects.

 P a g e | 229

11.66 This was a random block design with 5 treatment levels and 5 blocking
levels.

For both treatment and blocking effects, the critical value is F.05,4,16 =
3.01. The

observed F value for treatment effects is MSC / MSE = 35.98 / 7.36 =

4.89 which

is greater than the critical value. The null hypothesis for treatments is
rejected,

and we conclude that there is a significant different in treatment

means. No

multiple comparisons have been computed in the output. The

observed F value

for blocking effects is MSR / MSE = 10.36 /7.36 = 1.41 which is less than
the

critical value. There are no significant blocking effects. Using random

block

design on this experiment might have cost a loss of power.

11.67 This one-way ANOVA has 4 treatment levels and 24 observations. The
F = 3.51

yields a p-value of .034 indicating significance at  = .05. Since the

sample sizes

are equal, Tukey’s HSD is used to make multiple comparisons. The

computer
 P a g e | 230

output shows that means 1 and 3 are the only pairs that are
significantly different

(same signs in confidence interval). Observe on the graph that

the confidence intervals for means 1 and 3 barely overlap.

Chapter 12
Analysis of Categorical Data

LEARNING OBJECTIVES

This chapter presents several nonparametric statistics that can be used to

analyze data enabling you to:

1. Understand the chi-square goodness-of-fit test and how to use it.

2. Analyze data using the chi-square test of independence.

 P a g e | 231

CHAPTER TEACHING STRATEGY

Chapter 12 is a chapter containing the two most prevalent chi-square

tests: chi-square goodness-of-fit and chi-square test of independence. These
two techniques are important because they give the statistician a tool that is
particularly useful for analyzing nominal data (even though independent
variable categories can sometimes have ordinal or higher categories). It
should be emphasized that there are many instances in business research
where the resulting data gathered are merely categorical identification. For
example, in segmenting the market place (consumers or industrial users),
information is gathered regarding gender, income level, geographical
location, political affiliation, religious preference, ethnicity, occupation, size of
company, type of industry, etc. On these variables, the measurement is
often a tallying of the frequency of occurrence of individuals, items, or
companies in each category. The subject of the research is given no "score"
or "measurement" other than a 0/1 for being a member or not of a given
category. These two chi-square tests are perfectly tailored to analyze such
data.
 P a g e | 232

The chi-square goodness-of-fit test examines the categories of one

variable to determine if the distribution of observed occurrences matches
some expected or theoretical distribution of occurrences. It can be used to
determine if some standard or previously known distribution of proportions is
the same as some observed distribution of proportions. It can also be used to
validate the theoretical distribution of occurrences of phenomena such as
random arrivals that are often assumed to be Poisson distributed. You will
note that the degrees of freedom, k - 1 for a given set of expected values or
for the uniform distribution, change to k - 2 for an expected Poisson
distribution and to k - 3 for an expected normal distribution. To conduct a chi-
square goodness-of-fit test to analyze an expected Poisson distribution, the
value of lambda must be estimated from the observed data. This causes the
loss of an additional degree of freedom. With the normal distribution, both
the mean and standard deviation of the expected distribution are estimated
from the observed values causing the loss of two additional degrees of
freedom from the k - 1 value.

The chi-square test of independence is used to compare the observed

frequencies along the categories of two independent variables to expected
values to determine if the two variables are independent or not. Of course, if
the variables are not independent, they are dependent or related. This allows
business researchers to reach some conclusions about such questions as: is
smoking independent of gender or is type of housing preferred independent
of geographic region? The chi-square test of independence is often used as a
tool for preliminary analysis of data gathered in exploratory research where
the researcher has little idea of what variables seem to be related to what
variables, and the data are nominal. This test is particularly useful with
demographic type data.

A word of warning is appropriate here. When an expected frequency is

small, the observed chi-square value can be inordinately large thus yielding
an increased possibility of committing a Type I error. The research on this
problem has yielded varying results with some authors indicating that
expected values as low as two or three are acceptable and other researchers
demanding that expected values be ten or more. In this text, we have settled
on the fairly widespread accepted criterion of five or more.
 P a g e | 233

CHAPTER OUTLINE

12.1 Chi-Square Goodness-of-Fit Test

Testing a Population Proportion Using the Chi-square Goodness-

of-Fit

Test as an Alternative Technique to the z Test

12.2 Contingency Analysis: Chi-Square Test of Independence

KEY TERMS

Categorical Data Chi-Square Test of

Independence

Chi-Square Distribution Contingency Analysis

Chi-Square Goodness-of-Fit Test Contingency Table

 P a g e | 234

SOLUTIONS TO THE ODD-NUMBERED PROBLEMS IN CHAPTER 12

( fo  fe )2
fe
12.1 f0 fe

53 68 3.309

37 42 0.595
32 33 0.030
28 22 1.636

18 10 6.400

15 8 6.125

Ho: The observed distribution is the same

as the expected distribution.

Ha: The observed distribution is not the same

as the expected distribution.

( f0  fe )2
 
2

fe
Observed = 18.095

df = k - 1 = 6 - 1 = 5,  = .05

2.05,5 = 11.0705
 P a g e | 235

Since the observed 2 = 18.095 > 2.05,5 = 11.0705, the decision is to

reject the

null hypothesis.

The observed frequencies are not distributed the same as the expected

frequencies.
 P a g e | 236

( fo  fe )2
fe
12.2 f0 fe
19 18 0.056

17 18 0.056
14 18 0.889

18 18 0.000

19 18 0.056
21 18 0.500

18 18 0.000

18 18 0.000

fo = 144 fe = 144 1.557

Ho: The observed frequencies are uniformly distributed.

Ha: The observed frequencies are not uniformly distributed.

x
f 0

144
k 8
= 18

In this uniform distribution, each fe = 18

df = k – 1 = 8 – 1 = 7,  = .01

2.01,7 = 18.4753
 P a g e | 237

( f0  fe )2
 
2

fe
Observed = 1.557

Since the observed 2 = 1.557 < 2.01,7 = 18.4753, the decision is to

fail to reject

the null hypothesis

There is no reason to conclude that the frequencies are not

uniformly

distributed.
 P a g e | 238

12.3 Number f0 (Number)(f0)

0 28 0

1 17 17

2 11 22

3 _5 15

61
54

Ho: The frequency distribution is Poisson.

Ha: The frequency distribution is not Poisson.

54
61
= =0.9

Expected Expected

Number Probability Frequency

0 .4066 24.803

1 .3659 22.320

2 .1647 10.047

>3 .0628 3.831

Since fe for > 3 is less than 5, collapse categories 2 and >3:

( f o  fe )2
fe
Number fo fe

0 28 24.803 0.412
 P a g e | 239

1 17 22.320 1.268

>2 16 13.878 0.324

61 60.993 2.004

df = k - 2 = 3 - 2 = 1,  = .05

2.05,1 = 3.8415

( f0  fe )2
2  
fe
Observed = 2.001

Since the observed 2 = 2.001 < 2.05,1 = 3.8415, the decision is to fail
to reject

the null hypothesis.

There is insufficient evidence to reject the distribution as

Poisson distributed. The conclusion is that the distribution is
Poisson distributed.
 P a g e | 240

12.4

Category f(observed) Midpt. fm fm2

10-20 6 15 90 1,350

20-30 14 25 350 8,750

30-40 29 35 1,015 35,525

40-50 38 45 1,710 76,950

50-60 25 55 1,375 75,625

60-70 10 65 650 42,250

70-80 7 75 525 39,375

n = f = 129 fm = 5,715 fm2 = 279,825

x
 fm  5,715
 f 129
= 44.3

( fM ) 2 (5,715) 2
 fM 2  n
279,825 
129

n 1 128
s= = 14.43

Ho: The observed frequencies are normally distributed.

Ha: The observed frequencies are not normally distributed.

For Category 10 - 20 Prob

 P a g e | 241

10  44.3
14.43
z = = -2.38 .4913

20  44.3
14.43
z = = -1.68 - .4535

Expected prob.: .0378

For Category 20-30 Prob

for x = 20, z = -1.68 .4535

30  44.3
14.43
z= = -0.99 -.3389

Expected prob: .1146

 P a g e | 242

For Category 30 - 40 Prob

for x = 30, z = -0.99 .3389

40  44.3
14.43
z = = -0.30 -.1179

Expected prob: .2210

For Category 40 - 50 Prob

for x = 40, z = -0.30 .1179

50  44.3
14.43
z = = 0.40 +.1554

Expected prob: .2733

For Category 50 - 60 Prob

60  44.3
14.43
z = = 1.09 .3621

for x = 50, z = 0.40 -.1554

Expected prob: .2067

For Category 60 - 70 Prob

 P a g e | 243

70  44.3
14.43
z = = 1.78 .4625

for x = 60, z = 1.09 -.3621

Expected prob: .1004

For Category 70 - 80 Prob

80  44.3
14.43
z = = 2.47 .4932

for x = 70, z = 1.78 -.4625

Expected prob: .0307

For x < 10:

Probability between 10 and the mean, 44.3, = (.0378 + .1145 + .2210

+ .1179) = .4913. Probability < 10 = .5000 - .4912 = .0087

For x > 80:

Probability between 80 and the mean, 44.3, = (.0307 + .1004 + .2067

+ .1554) =

.4932. Probability > 80 = .5000 - .4932 = .0068

 P a g e | 244

Category Prob expected frequency

< 10 .0087 .0087(129) = 1.12

10-20 .0378 .0378(129) = 4.88

20-30 .1146 14.78

30-40 .2210 28.51

40-50 .2733 35.26

50-60 .2067 26.66

60-70 .1004 12.95

70-80 .0307 3.96

> 80 .0068 0.88

Due to the small sizes of expected frequencies, category < 10 is folded

into 10-20

and >80 into 70-80.

( fo  fe )2
fe
Category fo fe

10-20 6 6.00 .000

20-30 14 14.78 .041

30-40 29 28.51 .008

40-50 38 35.26 .213

50-60 25 26.66 .103

60-70 10 12.95 .672

70-80 7 4.84 .964

 P a g e | 245

2.001

( f0  fe )2
2  
fe
Calculated = 2.001

df = k - 3 = 7 - 3 = 4,  = .05

2.05,4 = 9.4877

Since the observed 2 = 2.004 < 2.05,4 = 9.4877, the decision is to fail
to reject

the null hypothesis. There is not enough evidence to declare that

the observed

frequencies are not normally distributed.

( fo  fe )2
fe
12.5 Definition fo Exp.Prop. fe

Happiness 42 .39 227(.39)= 88.53 24.46

Sales/Profit 95 .12 227(.12)= 27.24 168.55

Helping Others 27 .18 40.86 4.70

Achievement/
 P a g e | 246

Challenge 63 .31 70.37

0.77

227 198.48

Ho: The observed frequencies are distributed the same as the

expected

frequencies.

Ha: The observed frequencies are not distributed the same as the
expected

frequencies.

Observed 2 = 198.48

df = k – 1 = 4 – 1 = 3,  = .05

2.05,3 = 7.8147

Since the observed 2 = 198.48 > 2.05,3 = 7.8147, the decision is to

reject the

null hypothesis.

The observed frequencies for men are not distributed the same
as the

expected frequencies which are based on the responses of

women.
P a g e | 247
 P a g e | 248

12.6 Age fo Prop. from survey fe

( fo  fe )2
fe

10-14 22 .09 (.09)(212)=19.08 0.45

15-19 50 .23 (.23)(212)=48.76 0.03

20-24 43 .22 46.64 0.28

25-29 29 .14 29.68 0.02

30-34 19 .10 21.20 0.23

> 35 49 .22 46.64 0.12

212 1.13

Ho: The distribution of observed frequencies is the same as the

distribution of

expected frequencies.

Ha: The distribution of observed frequencies is not the same as the

distribution of

expected frequencies.

 = .01, df = k - 1 = 6 - 1 = 5

2.01,5 = 15.0863

The observed 2 = 1.13

Since the observed 2 = 1.13 < 2.01,5 = 15.0863, the decision is to fail
to reject
 P a g e | 249

the null hypothesis.

There is not enough evidence to declare that the distribution of

observed

frequencies is different from the distribution of expected

frequencies.
 P a g e | 250

12.7 Age fo m fm fm2

10-20 16 15 240 3,600

20-30 44 25 1,100 27,500

30-40 61 35 2,135 74,725

40-50 56 45 2,520 113,400

50-60 35 55 1,925 105,875

60-70 19 65 1,235 80,275

231 fm = 9,155 fm2 = 405,375

x
 fM 
9,155
n 231
= 39.63

( fM ) 2 (9,155) 2
 fM 2  n
405,375 
231

n 1 230
s = = 13.6

Ho: The observed frequencies are normally distributed.

Ha: The observed frequencies are not normally distributed.

For Category 10-20 Prob

10  39.63
13.6
z = = -2.18 .4854

20  39.63
13.6
z = = -1.44 -.4251
 P a g e | 251

Expected prob. .0603

For Category 20-30 Prob

for x = 20, z = -1.44 .4251

30  39.63
13.6
z = = -0.71 -.2611

Expected prob. .1640

For Category 30-40 Prob

for x = 30, z = -0.71 .2611

40  39.63
13.6
z = = 0.03 +.0120

Expected prob. .2731

For Category 40-50 Prob

50  39.63
13.6
z = = 0.76 .2764
 P a g e | 252

for x = 40, z = 0.03 -.0120

Expected prob. .2644

For Category 50-60 Prob

60  39.63
13.6
z = = 1.50 .4332

for x = 50, z = 0.76 -.2764

Expected prob. .1568

For Category 60-70 Prob

70  39.63
13.6
z = = 2.23 .4871

for x = 60, z = 1.50 -.4332

Expected prob. .0539

 P a g e | 253

For < 10:

Probability between 10 and the mean = .0603 + .1640 + .2611 = .

4854

Probability < 10 = .5000 - .4854 = .0146

For > 70:

Probability between 70 and the mean = .0120 + .2644 + .1568 + .

0539 = .4871

Probability > 70 = .5000 - .4871 = .0129

 P a g e | 254

Age Probability fe

< 10 .0146 (.0146)(231) = 3.37

10-20 .0603 (.0603)(231) = 13.93

20-30 .1640 37.88

30-40 .2731 63.09

40-50 .2644 61.08

50-60 .1568 36.22

60-70 .0539 12.45

> 70 .0129 2.98

Categories < 10 and > 70 are less than 5.

Collapse the < 10 into 10-20 and > 70 into 60-70.

( fo  fe )2
fe
Age fo fe

10-20 16 17.30 0.10

20-30 44 37.88 0.99

30-40 61 63.09 0.07

40-50 56 61.08 0.42

50-60 35 36.22 0.04

60-70 19 15.43 0.83

2.45

df = k - 3 = 6 - 3 = 3,  = .05
 P a g e | 255

2.05,3 = 7.8147

Observed 2 = 2.45

Since the observed 2 < 2.05,3 = 7.8147, the decision is to fail to

reject the null

hypothesis.

There is no reason to reject that the observed frequencies are

normally

distributed.
 P a g e | 256

12.8 Number f ( f) (number)

0 18 0

1 28 28
2 47 94

3 21 63

4 16 64
5 11 55

6 or more 9 54

f = 150 f(number) = 358

 f  number  358
f 150
= = 2.4

Ho: The observed frequencies are Poisson distributed.

Ha: The observed frequencies are not Poisson distributed.

Number Probability fe

0 .0907 (.0907)(150) = 13.61

1 .2177 (.2177)(150) = 32.66

2 .2613 39.20

3 .2090 31.35

4 .1254 18.81

5 .0602 9.03

6 or more .0358 5.36

 P a g e | 257

( f0  fe )2
f0
fo fe

18 13.61 1.42

28 32.66 0.66

47 39.20 1.55

21 31.35 3.42
16 18.81 0.42

11 9.03 0.43

9 5.36 2.47

10.37

The observed 2 = 10.37

 = .01, df = k – 2 = 7 – 2 = 5, 2.01,5 = 15.0863

Since the observed 2 = 10.37 < 2.01,5 = 15.0863, the decision is to fail
to reject

the null hypothesis.

There is not enough evidence to reject the claim that the

observed

frequencies are Poisson distributed.

12.9 H0: p = .28 n = 270 x = 62

Ha: p  .28
 P a g e | 258

( fo  f e )2
fe
fo fe

Spend More 62 270(.28) = 75.6 2.44656

Don't Spend More 208 270(.72) = 194.4 0.95144

Total 270 270.0 3.39800

The observed value of 2 is 3.398

 = .05 and /2 = .025 df = k - 1 = 2 - 1 = 1

2.025,1 = 5.02389

Since the observed 2 = 3.398 < 2.025,1 = 5.02389, the decision is to

fail to

reject the null hypothesis.

12.10 H0: p = .30 n = 180 x= 42

Ha: p < .30

 P a g e | 259

( fo  fe )2
fe
f0 fe

Provide 42 180(.30) = 54 2.6666

Don't Provide 138 180(.70) = 126 1.1429

Total 180 180 3.8095

The observed value of 2 is 3.8095

 = .05 df = k - 1 = 2 - 1 = 1

2.05,1 = 3.8415

Since the observed 2 = 3.8095 < 2.05,1 = 3.8415, the decision is to

fail to

reject the null hypothesis.

 P a g e | 260

12.11

Variable Two

Variabl 203 326 529

e
68 110 178
One
271 707
436

Ho: Variable One is independent of Variable Two.

Ha: Variable One is not independent of Variable Two.

(529)( 271) (529)( 436)

707 707
e11 = = 202.77 e12 = = 326.23

(271)(178) (436)(178)
707 707
e21 = = 68.23 e22 = = 109.77

Variable Two

Variabl (202.7 (326.2

e 7) 3)
529
One 203 326

(109.7
(68.23) 7) 178

68 110
 P a g e | 261

271 707
436

(203  202.77) 2 (326  326.23) 2 (68  6.23) 2

202.77 326.23 68.23
2 = + + +

(110  109.77) 2
109.77
=

.00 + .00 + .00 + .00 = 0.00

 = .01, df = (c-1)(r-1) = (2-1)(2-1) = 1

2.01,1 = 6.6349

Since the observed 2 = 0.00 < 2.01,1 = 6.6349, the decision is to fail
to reject

the null hypothesis.

Variable One is independent of Variable Two.

 P a g e | 262

12.12

Variable

Two

Variabl 24 13 47 58
e 142
93 59 187 244
One
583

117 72 234
302 725

Ho: Variable One is independent of Variable Two.

Ha: Variable One is not independent of Variable Two.

(142)(117) (142)( 72)

725 725
e11 = = 22.92 e12 = = 14.10

(142)( 234) (142)(302)

725 725
e13 = = 45.83 e14 = = 59.15

(583)(117) (583)(72)
725 725
e21 = = 94.08 e22 = = 57.90

(583)( 234) (583)(302)

725 725
e23 = = 188.17 e24 = = 242.85
 P a g e | 263

Variable Two

Variabl (22.9 (14.1 (45.83) (59.15)

e 2) 0)
47 58
One 24 13 142

(94.0 (57.9 (188.1 (242.85

8) 0) 7) )

93 59 187 244 583

117 72 234
302 725

(24  22.92) 2 (13  14.10) 2 (47  45.83) 2 (58  59.15) 2

22.92 14.10 45.83 59.15
2 = + + + +

(93  94.08) 2 (59  57.90) 2 (188  188.17) 2 (244  242.85) 2

94.08 57.90 188.17 242.85
+ + +
=

.05 + .09 + .03 + .02 + .01 + .02 + .01 + .01 = 0.24

 = .01, df = (c-1)(r-1) = (4-1)(2-1) = 3, 2.01,3 = 11.3449

Since the observed 2 = 0.24 < 2.01,3 = 11.3449, the decision is to

fail to

reject the null hypothesis.

Variable One is independent of Variable Two.

 P a g e | 264

12.13

Social Class

Lower Middle
Upper
Numbe
r 0 7 18 6 31

of 1 9 38 23 70

Childre 2 or 34 97 58 189
n 3
47 31 30 108
>3

97 184 398
117

Ho: Social Class is independent of Number of Children.

Ha: Social Class is not independent of Number of Children.

(31)(97) (189)(97)
398 398
e11 = = 7.56 e31 = = 46.06

(31)(184) (189)(184)
398 398
e12 = = 14.3 e32 = = 87.38

(31)(117) (189)(117)
398 398
e13 = = 9.11 e33 = = 55.56
 P a g e | 265

(70)(97) (108)(97)
398 398
e21 = = 17.06 e41 = = 26.32

(70)(184) (108)(184)
398 398
e22 = = 32.36 e42 = = 49.93

(70)(117) (108)(117)
398 398
e23 = = 20.58 e43 = = 31.75

Social Class

Lower Middle
Upper

0 (7.56) (14.3 (9.11

Numbe 3) )
r 7 31
18 6
of 1
(17.0 (32.3 (20.5
Childre 6) 6) 8) 70
n
2 or 9 38 23
3
(46.0 (87.3 (55.5 189
6) 8) 6)

>3 34 97 58
108
(26.3 (49.9 (31.7
2) 3) 5)

47 31 30

97 184 398
117
 P a g e | 266

(7  7.56) 2 (18  14.33) 2 (6  9.11) 2 (9  17.06) 2

7.56 14.33 9.11 17.06
2 = + + + +

(38  32.36) 2 (23  20.58) 2 (34  46.06) 2 (97  87.38) 2

32.36 20.58 46.06 87.38
+ + + +

(58  55.56) 2 (47  26.32) 2 (31  49.93) 2 (30  31.75) 2

55.56 26.32 49.93 31.75
+ + + =

.04 + .94 + 1.06 + 3.81 + .98 + .28 + 3.16 + 1.06 + .11 + 16.25
+

7.18 + .10 = 34.97

 = .05, df = (c-1)(r-1) = (3-1)(4-1) = 6

2.05,6 = 12.5916

Since the observed 2 = 34.97 > 2.05,6 = 12.5916, the decision is to

reject the

null hypothesis.

Number of children is not independent of social class.

 P a g e | 267

12.14

Type of Music Preferred

Rock R&B Coun Clssic

NE 140 32 5 18 195

Regio S 134 41 52 8 235

n
W 154 27 8 13 202

632

428 100 65
39

Ho: Type of music preferred is independent of region.

Ha: Type of music preferred is not independent of region.

(195)( 428) (235)(65)

632 632
e11 = = 132.6 e23 = = 24.17

(195)(100) (235)(39)
632 632
e12 = = 30.85 e24 = = 14.50
 P a g e | 268

(195)(65) (202)( 428)

632 632
e13 = = 20.06 e31 = = 136.80

(195)(39) (202)(100)
632 632
e14 = = 12.03 e32 = = 31.96

(235)( 428) (202)( 65)

632 632
e21 = = 159.15 e33 = = 20.78

(235)(100) (202)(39)
632 632
e22 = = 37.18 e34 = = 12.47

Type of Music Preferred

Rock R&B Coun Clssic

NE (132.0 (30.8 (20.0 (12.0 195

6) 5) 6) 3)
Regio 235
n 140 32 5 18
202
S (159.1 (37.1 (24.1 (14.5
5) 8) 7) 0) 632

134 41 52 8

W (136.8 (31.9 (20.7 (12.4

0) 6) 8) 7)

154 27 8 13
 P a g e | 269

428 100 65
39

(141  132.06) 2 (32  30.85) 2 (5  20.06) 2 (18  12.03) 2

132.06 30.85 20.06 12.03
2 = + + + +

(134  159.15) 2 (41  37.18) 2 (52  24.17) 2 (8  14.50) 2

159.15 37.18 24.17 14.50
+ + + +

(154  136.80) 2 (27  31.96) 2 (8  20.78) 2 (13  12.47) 2

136.80 31.96 20.78 12.47
+ + + =

.48 + .04 + 11.31 + 2.96 + 3.97 + .39 + 32.04 + 2.91 + 2.16

+ .77 +

7.86 + .02 = 64.91

 = .01, df = (c-1)(r-1) = (4-1)(3-1) = 6

2.01,6 = 16.8119

Since the observed 2 = 64.91 > 2.01,6 = 16.8119, the decision is to

reject the null hypothesis.

Type of music preferred is not independent of region of the country.

12.15
 P a g e | 270

Transportation Mode

Air Train Truck

Industr Publishing 32 12 41 85
y
Comp.Har 5 6 24 35
d.
120
37 18 65

H0: Transportation Mode is independent of Industry.

Ha: Transportation Mode is not independent of Industry.

(85)(37) (35)(37)
120 120
e11 = = 26.21 e21 = = 10.79

(85)(18) (35)(18)
120 120
e12 = = 12.75 e22 = = 5.25

(85)(65) (35)(65)
120 120
e13 = = 46.04 e23 = = 18.96

Transportation Mode

Air Train Truck

Industr Publishing (26.2 (12.7 (46.0

y 1) 5) 4)
85
32 12 41
 P a g e | 271

Comp.Har (10.7 (5.25 (18.9

d. 9) ) 6)
35
5 6 24
120
37 18 65

(32  26.21) 2 (12  12.75) 2 (41  46.04) 2

26.21 12.75 46.04
2 = + + +

(5  10.79) 2 (6  5.25) 2 (24  18.96) 2

10.79 5.25 18.96
+ + =

1.28 + .04 + .55 + 3.11 + .11 + 1.34 = 6.43

 = .05, df = (c-1)(r-1) = (3-1)(2-1) = 2

2.05,2 = 5.9915

Since the observed 2 = 6.43 > 2.05,2 = 5.9915, the decision is to

reject the null hypothesis.

Transportation mode is not independent of industry.

12.16
 P a g e | 272

Number of
Bedrooms

<2 3 >4

Number 1 116 101 57 274

of
2 90 325 160 575
Stories
206 426 849
217

H0: Number of Stories is independent of number of bedrooms.

Ha: Number of Stories is not independent of number of bedrooms.

(274)( 206) (575)( 206)

849 849
e11 = = 66.48 e21 = = 139.52

(274)( 426) (575)( 426)

849 849
e12 = = 137.48 e22 = = 288.52

(274)( 217) (575)( 217)

849 849
e13 = = 70.03 e23 = = 146.97

(90  139.52) 2 (101  137.48) 2 (57  70.03) 2 (90  139.52) 2

139.52 137.48 70.03 139.52
2 = + + +
+
 P a g e | 273

(325  288.52) 2 (160  146.97) 2

288.52 146.97
+ =

2 = 36.89 + 9.68 + 2.42 + 17.58 + 4.61 + 1.16 = 72.34

 = .10 df = (c-1)(r-1) = (3-1)(2-1) = 2

2.10,2 = 4.6052

Since the observed 2 = 72.34 > 2.10,2 = 4.6052, the decision is to

reject the null hypothesis.

Number of stories is not independent of number of bedrooms.

 P a g e | 274

12.17

Mexican Citizens

Yes No

Type Dept. 24 17 41

of Disc. 20 15 35
Hard. 11 19 30
Store Shoe 32 28
60

87 79 166

Ho: Citizenship is independent of store type

Ha: Citizenship is not independent of store type

(41)(87) (30)(87)
166 166
e11 = = 21.49 e31 = = 15.72

(41)(79) (30)(79)
166 166
e12 = = 19.51 e32 = = 14.28

(35)(87) (60)(87)
166 166
e21 = = 18.34 e41 = = 31.45

(35)(79) (60)(79)
166 166
e22 = = 16.66 e42 = = 28.55
 P a g e | 275

Mexican Citizens

Yes No

Type Dept. (21.4 (19.5

9) 1)
of 41
24 17

Store Disc. (18.3 (16.6

4) 6) 35

20 15

Hard. (15.7 (14.2 30

2) 8)

11 19
60
Shoe (31.4 (28.5
5) 5)

32 28

87 79 166

(24  21.49) 2 (17  19.51) 2 (20  18.34) 2 (15  16.66) 2

21.49 19.51 18.34 16.66
2 = + + + +

(11  15.72) 2 (19  14.28) 2 (32  31.45) 2 (28  28.55) 2

15.72 14.28 31.45 28.55
+ + + =

.29 + .32 + .15 + .17 + 1.42 + 1.56 + .01 + .01 =

3.93
 P a g e | 276

 = .05, df = (c-1)(r-1) = (2-1)(4-1) = 3

2.05,3 = 7.8147

Since the observed 2 = 3.93 < 2.05,3 = 7.8147, the decision is to fail
to

reject the null hypothesis.

Citizenship is independent of type of store.

12.18  = .01, k = 7, df = 6

H0: The observed distribution is the same as the expected distribution

Ha: The observed distribution is not the same as the expected

distribution

Use:

( f0  fe )2
2  
fe

critical 2.01,6 = 16.8119

 P a g e | 277

( fo  fe )2
fe
fo fe (f0-fe)2

214 206 64 0.311

235 232 9 0.039

279 268 121 0.451

281 284 9 0.032

264 268 16 0.060

254 232 484 2.086

211 206 25 0.121

3.100

( f0  fe )2
2  
fe
= 3.100

Since the observed value of 2 = 3.1 < 2.01,6 = 16.8119, the decision
is to fail to

reject the null hypothesis. The observed distribution is not different

from the

expected distribution.

12.19

Variable 2

12 23 21 56
 P a g e | 278

Variable 8 17 20 45
1
7 11 18 36

27 51 59 137

e11 = 11.04 e12 = 20.85 e13 = 24.12

e21 = 8.87 e22 = 16.75 e23 = 19.38

e31 = 7.09 e32 = 13.40 e33 = 15.50

(12  11.04) 2 (23  20.85) 2 (21  24.12) 2 (8  8.87) 2

11.04 20.85 24.12 8.87
2 = + + + +

(17  16.75) 2 (20  19.38) 2 (7  7.09) 2 (11  13.40) 2

16.75 19.38 7.09 13.40
+ + + +

(18  15.50) 2
15.50
=

.084 + .222 + .403 + .085 + .004 + .020 + .001 + .430 + .402

= 1.652

df = (c-1)(r-1) = (2)(2) = 4  = .05

2.05,4 = 9.4877
 P a g e | 279

Since the observed value of 2 = 1.652 < 2.05,4 = 9.4877, the

decision is to fail

to reject the null hypothesis.

 P a g e | 280

12.20

Location

NE W S

Custom Industri 230 115 68 413

er al

Retail 185 143 89 417

415 258 157 830

(413)( 415) (417)( 415)

830 830
e11 = = 206.5 e21 = = 208.5

(413)(258) (417)( 258)

830 830
e12 = = 128.38 e22 = = 129.62

(413)(157) (417)(157)
830 830
e13 = = 78.12 e23 = = 78.88

Location

NE W S

Custom Industri (206. (128.3 (78.1

er al 5) 8) 2)
413
230 115 68
 P a g e | 281

Retail (208. (129.6 (78.8

5) 2) 8)
417
185 143 89

415 258 157 830

(230  206.5) 2 (115  128.38) 2 (68  78.12) 2

206.5 128.38 78.12
2 = + + +

(185  208.5) 2 (143  129.62) 2 (89  78.88) 2

208.5 129.62 78.88
+ + =

2.67 + 1.39 + 1.31 + 2.65 + 1.38 + 1.30 = 10.70

 = .10 and df = (c - 1)(r - 1) = (3 - 1)(2 - 1) = 2

2.10,2 = 4.6052

Since the observed 2 = 10.70 > 2.10,2 = 4.6052, the decision is to

reject the

null hypothesis.

Type of customer is not independent of geographic region.

12.21 Cookie Type fo

Chocolate Chip 189

 P a g e | 282

Peanut Butter 168

Cheese Cracker 155

Lemon Flavored 161

Chocolate Mint 216

Vanilla Filled 165

fo = 1,054

Ho: Cookie Sales is uniformly distributed across kind of cookie.

Ha: Cookie Sales is not uniformly distributed across kind of cookie.

f 0

1,054
no.kinds 6
If cookie sales are uniformly distributed, then fe = =
175.67

( fo  fe )2
fe
fo fe

189 175.67 1.01

168 175.67 0.33

155 175.67 2.43

161 175.67 1.23

216 175.67 9.26

165 175.67 0.65

14.91
 P a g e | 283

The observed 2 = 14.91

 = .05 df = k - 1 = 6 - 1 = 5

2.05,5 = 11.0705

Since the observed 2 = 14.91 > 2.05,5 = 11.0705, the decision is to

reject the

null hypothesis.

Cookie Sales is not uniformly distributed by kind of cookie.

 P a g e | 284

12.22

Gender

M F

Boug Y 207 65 272

ht
N 811 984 1,795
Car

1,018 1,04 2,067

9

Ho: Purchasing a car or not is independent of gender.

Ha: Purchasing a car or not is not independent of gender.

(272)(1,018) (27)(1,049)
2,067 2,067
e11 = = 133.96 e12 = = 138.04

(1,795)(1,018) (1,795)(1,049)
2,067 2,067
e21 = = 884.04 e22 = = 910.96

Gender

M F

Boug Y (133.9 (138.0

ht 6) 4)
272
Car 207 65

N (884.0 (910.9
4) 6)
1,795
811 984
 P a g e | 285

1,018 1,049 2,067

(207  133.96) 2 (65  138.04) 2 (811  884.04) 2

133.96 138.04 884.04
2 = + + +

(984  910.96) 2
910.96
= 39.82 + 38.65 + 6.03 + 5.86 = 90.36

 = .05 df = (c-1)(r-1) = (2-1)(2-1) = 1

2.05,1 = 3.8415

Since the observed 2 = 90.36 > 2.05,1 = 3.8415, the decision is to

reject the

null hypothesis.

Purchasing a car is not independent of gender.

 P a g e | 286

12.23 Arrivals fo ( fo)(Arrivals)

0 26 0

1 40 40

2 57 114

3 32 96

4 17 68

5 12 60

6 8 48

fo = 192 (fo)(arrivals) = 426

 ( f )(arrivals )  426
0

f 0 192
 = = 2.2

Ho: The observed frequencies are Poisson distributed.

Ha: The observed frequencies are not Poisson distributed.

Arrivals Probability fe

0 .1108 (.1108)(192) = 21.27

1 .2438 (.2438)(192) = 46.81

2 .2681 51.48

3 .1966 37.75

4 .1082 20.77

5 .0476 9.14

6 .0249 4.78
 P a g e | 287

( fo  fe )2
fe
fo fe

26 21.27 1.05

40 46.81 0.99

57 51.48 0.59

32 37.75 0.88

17 20.77 0.68

12 9.14 0.89

8 4.78 2.17

7.25

Observed 2 = 7.25

 = .05 df = k - 2 = 7 - 2 = 5

2.05,5 = 11.0705

Since the observed 2 = 7.25 < 2.05,5 = 11.0705, the decision is to fail
to reject

the null hypothesis. There is not enough evidence to reject the

claim that the

observed frequency of arrivals is Poisson distributed.

 P a g e | 288

12.24 Ho: The distribution of observed frequencies is the same as the

distribution of expected frequencies.

Ha: The distribution of observed frequencies is not the same as the

distribution of

expected frequencies.

Soft Drink fo proportions fe

( fo  fe )2
fe

Classic Coke 314 .179 (.179)(1726) = 308.95 0.08

Pepsi 219 .115 (.115)(1726) =

198.49 2.12

Diet Coke 212 .097 167.42 11.87

Mt. Dew 121 .063 108.74 1.38

Diet Pepsi 98 .061

105.29 0.50

Sprite 93 .057 98.32 0.29

Dr. Pepper 88 .056 96.66 0.78

Others 581 .372 642.07 5.81

fo = 1,726
22.83
 P a g e | 289

Observed 2 = 22.83

 = .05 df = k - 1 = 8 - 1 = 7

2.05,7 = 14.0671

Since the observed 2 = 22.83 > 2.05,6 = 14.0671, the decision is to

reject the

null hypothesis.

The observed frequencies are not distributed the same as the expected
frequencies

from the national poll.

 P a g e | 290

12.25

Position

Syste
ms
Manag Programm Operat Analys
er er or t

0-3 6 37 11 13 67

4-8 28 16 23 24 91
Years
>8 47 10 12 19 88

81 63 46 56 246

(67)(81) (91)( 46)

246 246
e11 = = 22.06 e23 = = 17.02

(67)(63) (91)(56)
246 246
e12 = = 17.16 e24 = = 20.72

(67)( 46) (88)(81)

246 246
e13 = = 12.53 e31 = = 28.98

(67)(56) (88)(63)
246 246
e14 = = 15.25 e32 = = 22.54
 P a g e | 291

(91)(81) (88)( 46)

246 246
e21 = = 29.96 e33 = = 16.46

(91)(63) (88)(56)
246 246
e22 = = 23.30 e34 = = 20.03

Position

Syste
ms
Manag Programm Operat Analys
er er or t

0-3 (17.16) 67
(22.06) (12.53) (15.25
37 )
Years 6 11
13

4-8 (23.30) 91
(29.96) (17.02) (20.72
16 )
28 23
24

>8 (22.54) 88
(28.98) (16.46) (20.03
10 )
47 12
19

81 63 46 56 246
 P a g e | 292

(6  22.06) 2 (37  17.16) 2 (11  12.53) 2 (13  15.25) 2

22.06 17.16 12.53 15.25
2 = + + + +

(28  29.96) 2 (16  23.30) 2 (23  17.02) 2 (24  20.72) 2

29.96 23.30 17.02 20.72
+ + + +

(47  28.98) 2 (10  22.54) 2 (12  16.46)2 (19  20.03) 2

28.98 22.54 16.46 20.03
+ + + =

11.69 + 22.94 + .19 + .33 + .13 + 2.29 + 2.1 + .52 + 11.2 +

6.98 +

1.21 + .05 = 59.63

 = .01 df = (c-1)(r-1) = (4-1)(3-1) = 6

2.01,6 = 16.8119

Since the observed 2 = 59.63 > 2.01,6 = 16.8119, the decision is to

reject the

null hypothesis. Position is not independent of number of years of

experience.
 P a g e | 293

12.26 H0: p = .43 n = 315  =.05

Ha: p  .43 x = 120 /2 = .025

( fo  fe )2
fe
fo fe

More Work,

More Business 120 (.43)(315) = 135.45 1.76

Others 195 (.57)(315) = 179.55 1.33

Total 315 315.00 3.09

The observed value of 2 is 3.09

 = .05 and /2 = .025 df = k - 1 = 2 - 1 = 1

2.025,1 = 5.0239

Since 2 = 3.09 < 2.025,1 = 5.0239, the decision is to fail to reject

the null

hypothesis.

12.27
 P a g e | 294

Type of College or
University

Communi Large Small

ty
Universit Colleg
College y e

Numbe 0 25 178 31 234

r
1 49 141 12 202
of
2 31 54 8 93
Childre
>3 22 14 6 42
n

127 387 57 571

Ho: Number of Children is independent of Type of College or

University.

Ha: Number of Children is not independent of Type of College or

University.

(234)(127) (93)(127)
571 571
e11 = = 52.05 e31 = = 20.68

(234)(387) (193)(387)
571 571
e12 = = 158.60 e32 = = 63.03

(234)(57) (93)(57)
571 571
e13 = = 23.36 e33 = = 9.28
 P a g e | 295

(202)(127) (42)(127)
571 571
e21 = = 44.93 e41 = = 9.34

(202)(387) (42)(387)
571 571
e22 = = 136.91 e42 = = 28.47

(202)(57) (42)(57)
571 571
e23 = = 20.16 e43 = = 4.19

Type of College or
University

Communi Large Small

ty
Universit Colleg
College y e

Numbe 0 (52.05) (23.36 234

r (158.60) )
25
of 178 31

Childre 1 (44.93) (20.16 202

n (136.91) )
49
141 12

2 (20.68) (63.03) (9.28) 93

31 54 8

>3 (9.34) (28.47) (4.19) 42

22 14 6

127 387 57 571

 P a g e | 296

(25  52.05) 2 (178  158.6) 2 (31  23.36) 2 (49  44.93) 2

52.05 158.6 23.36 44.93
2 = + + + +

(141  136.91) 2 (12  20.16) 2 (31  20.68) 2 (54  63.03) 2

136.91 20.16 20.68 63.03
+ + + +

(8  9.28) 2 (22  9.34) 2 (14  28.47 ) 2 (6  4.19) 2

9.28 9.34 28.47 4.19
+ + + =

14.06 + 2.37 + 2.50 + 0.37 + 0.12 + 3.30 + 5.15 + 1.29 + 0.18

+

17.16 + 7.35 + 0.78 = 54.63

 = .05, df= (c - 1)(r - 1) = (3 - 1)(4 - 1) = 6

2.05,6 = 12.5916

Since the observed 2 = 54.63 > 2.05,6 = 12.5916, the decision is to

reject the

null hypothesis.

Number of children is not independent of type of College or University.

 P a g e | 297

12.28 The observed chi-square is 30.18 with a p-value of .0000043. The chi-
square
goodness-of-fit test indicates that there is a significant difference
between the observed frequencies and the expected frequencies. The
distribution of responses to the question is not the same for adults
between 21 and 30 years of age as they are for others. Marketing and
sales people might reorient their 21 to 30 year old efforts away from
home improvement and pay more attention to leisure travel/vacation,
clothing, and home entertainment.

12.29 The observed chi-square value for this test of independence is 5.366.
The associated p-value of .252 indicates failure to reject the null
hypothesis. There is not enough evidence here to say that color choice
is dependent upon gender. Automobile marketing people do not have
to worry about which colors especially appeal to men or to women
because car color is independent of gender. In addition, design and
production people can determine car color quotas based on other
variables.

Chapter 13
 P a g e | 298

Nonparametric Statistics

LEARNING OBJECTIVES

This chapter presents several nonparametric statistics that can be used to

analyze data enabling you to:

1. Recognize the advantages and disadvantages of nonparametric

statistics.

2. Understand how to use the runs test to test for randomness.

3. Know when and how to use the Mann-Whitney U Test, the Wilcoxon
matched-pairs signed rank test, the Kruskal-Wallis test, and the
Friedman test.

4. Learn when and how to measure correlation using Spearman's rank

correlation measurement.

CHAPTER TEACHING STRATEGY

Chapter 13 contains new six techniques for analysis. Only the first
technique, the runs test, is conceptually a different idea for the student to
consider than anything presented in the text to this point. The runs test is a
mechanism for testing to determine if a string of data are random. There is a
 P a g e | 299

runs test for small samples that uses Table A.12 in the appendix and a test for
large samples, which utilizes a z test.

The main portion of chapter 13 (middle part) contains nonparametric alternatives to

parametric tests presented earlier in the book. The Mann-Whitney U test is a
nonparametric alternative to the t test for independent means. The Wilcoxon matched-
pairs signed ranks test is an alternative to the t test for matched-pairs. The Kruskal-Wallis
is a nonparametric alternative to the one-way analysis of variance test. The Friedman test
is a nonparametric alternative to the randomized block design presented in chapter 11.
Each of these four tests utilizes rank analysis.

The last part of the chapter is a section on Spearman's rank

correlation. This correlation coefficient can be presented as a nonparametric
alternative to the Pearson product-moment correlation coefficient of chapter
3. Spearman's rank correlation uses either ranked data or data that is
converted to ranks. The interpretation of Spearman's rank correlation is
similar to Pearson's product-moment correlation coefficient.

CHAPTER OUTLINE
 P a g e | 300

13.1 Runs Test

Small-Sample Runs Test

Large-Sample Runs Test

13.2 Mann-Whitney U Test

Small-Sample Case
Large-Sample Case

13.3 Wilcoxon Matched-Pairs Signed Rank Test

Small-Sample Case (n < 15)

Large-Sample Case (n > 15)

13.4 Kruskal-Wallis Test

13.5 Friedman Test

13.6 Spearman's Rank Correlation

KEY TERMS

Friedman Test Parametric Statistics

Kruskal-Wallis Test Runs Test

Mann-Whitney U Test Spearman’s Rank Correlation

 P a g e | 301

Nonparametric Statistics
Wilcoxon Matched-Pairs Signed Rank Test
 P a g e | 302

SOLUTIONS TO CHAPTER 13

13.1 Ho: The observations in the sample are randomly generated.

Ha: The observations in the sample are not randomly generated.

This is a small sample runs test since n1, n2 < 20

 = .05, The lower tail critical value is 6 and the upper tail critical value
is 16

n1 = 10 n2 = 10

R = 11

Since R = 11 is between the two critical values, the decision is to fail

to reject the null hypothesis.

The data are random.

13.2 Ho: The observations in the sample are randomly generated.

Ha: The observations in the sample are not randomly generated.

 P a g e | 303

 = .05, /2 = .025, z.025= + 1.96

n1 = 26 n2 = 21 n = 47

2n1n2 2(26)( 21)

R  1  1
n1  n2 26  21
= 24.234

2n1n2 (2n1n2  n1  n2 ) 2(26)( 21) 2(26)( 21)  26  21

R  
(n1  n2 ) 2 (n1  n2  1) (26  21) 2 (26  21  1)
= 3.351

R=9

R   R 9  24.234
z 
R 3.351
= -4.55

Since the observed value of z = -4.55 < z.025 = -1.96, the decision is to
reject the

null hypothesis. The data are not randomly generated.

 P a g e | 304

13.3 n1 = 8 n2 = 52  = .05

This is a two-tailed test and /2 = .025. The p-value from the
printout is .0264.

Since the p-value is the lowest value of “alpha” for which the null
hypothesis can

be rejected, the decision is to fail to reject the null hypothesis

(p-value = .0264 > .025). There is not enough evidence to reject that
the data are randomly generated.

13.4 The observed number of runs is 18. The mean or expected

number of runs

is 14.333. The p-value for this test is .1452. Thus, the test is not
significant at alpha of .05 or .025 for a two-tailed test. The decision is to
fail to reject the

null hypothesis. There is not enough evidence to declare that the

data are not random. Therefore, we must conclude that the data a
randomly generated.

13.5 Ho: The observations in the sample are randomly generated.

Ha: The observations in the sample are not randomly generated.

 P a g e | 305

Since n1, n2 > 20, use large sample runs test

 = .05 Since this is a two-tailed test, /2 = .025, z.025 = + 1.96. If

the

observed value of z is greater than 1.96 or less than -1.96, the decision
is to reject the null hypothesis.

R = 27 n1 = 40 n2 = 24

2n1n2 2(40)( 24)

R  1  1
n1  n2 64
= 31

2n1n2 (2n1n2  n1  n2 ) 2(40)( 24) 2(40)( 24)  40  24

R  
(n1  n2 ) 2 (n1  n2  1) (64) 2 (63)
= 3.716

R   R 27  31
z 
R 3.716
= -1.08

Since the observed z of -1.08 is greater than the critical lower tail z
value

of -1.96, the decision is to fail to reject the null hypothesis. The

data are randomly generated.
 P a g e | 306

13.6 Ho: The observations in the sample are randomly generated.

Ha: The observations in the sample are not randomly generated.

n1 = 5 n2 = 8 n = 13  = .05

Since this is a two-tailed test, /2 = .025

From Table A.11, the lower critical value is 3

From Table A.11, the upper critical value is 11

R=4

Since R = 4 > than the lower critical value of 3 and less than the upper
critical

value of 11, the decision is to fail to reject the null hypothesis.

The data are randomly generated.

13.7 Ho: Group 1 is identical to Group 2

Ha: Group 1 is not identical to Group 2

Use the small sample Mann-Whitney U test since both n1, n2 < 10,  = .
05. Since this is a two-tailed test, /2 = .025. The p-value is obtained
using Table A.13.

Value Rank Group

11 1 1
 P a g e | 307

13 2.5 1

13 2.5 2
14 4 2

15 5 1

17 6 1
18 7.5 1

18 7.5 2

21 9.5 1

21 9.5 2

22 11 1

23 12.5 2

23 12.5 2

24 14 2

26 15 1

29 16 1

n1 = 8

n2 = 8

W1 = 1 + 2.5 + 5 + 6 + 7.5 + 9.5 + 15 + 16 = 62.5

n1 (n1  1) (8)(9)
U  n1  n2   W1  (8)(8)   62.5
2 2
= 37.5

U '  n1  n2  U
= 64 – 37.5 = 26.5
 P a g e | 308

We use the small U which is 26.5

From Table A.13, the p-value for U = 27 is .3227(2) = .6454

Since this p-value is greater than /2 = .025, the decision is to fail to
reject the

null hypothesis.

13.8 Ho: Population 1 has values that are no greater than population 2

Ha: Population 1 has values that are greater than population 2

Value Rank Group

203 1 2

208 2 2

209 3 2

211 4 2

214 5 2

216 6 1

217 7 1

218 8 2

219 9 2

222 10 1

223 11 2

224 12 1
 P a g e | 309

227 13 2

229 14 2

230 15.5 2

230 15.5 2

231 17 1

236 18 2

240 19 1

241 20 1

248 21 1

255 22 1

256 23 1

283 24 1

n1 = 11

n2 = 13

W1 = 6 + 7 + 10 + 12 + 17 + 19 + 20 + 21 + 22 + 23 + 24 =

W1 = 181

n1  n2 (11)(13)
 
2 2
= 71.5

n1  n2 (n1  n2  1) (11)(13)( 25)

  
12 12
= 17.26
 P a g e | 310

n1 (n1  1) (11)(12)
U  n1  n2   W1  (11)(13   181
2 2
= 28

U   28  71.5
z 
 17.26
= -2.52

 = .01, z.01 = 2.33

 2.52
Since z = = 2.52 > z = 2.33, the decision is to reject the
null

hypothesis.

13.9 Contacts Rank Group

6 1 1

8 2 1

9 3.5 1

9 3.5 2

10 5 2

11 6.5 1

11 6.5 1
 P a g e | 311

12 8.5 1

12 8.5 2

13 11 1

13 11 2

13 11 2

14 13 2

15 14 2

16 15 2

17 16 2

W1 = 39

n1 ( n1  1) (7)(8)
U1  n1  n2   W1  (7)(9)   39
2 2
= 52

U 2  n1  n2  U1
= (7)(9) – 52 = 11

U = 11

From Table A.13, the p-value = .0156. Since this p-value is greater
than  = .01,

the decision is to fail to reject the null hypothesis.

13.10 Ho: Urban and rural spend the same

 P a g e | 312

Ha: Urban and rural spend different amounts

Expenditure Rank Group

1950 1 U

2050 2 R

2075 3 R

2110 4 U

2175 5 U

2200 6 U

2480 7 U

2490 8 R

2540 9 U

2585 10 R

2630 11 U

2655 12 U

2685 13 R

2710 14 U

2750 15 U

2770 16 R

2790 17 R

2800 18 R

2850 19.5 U

2850 19.5 U

2975 21 R

2995 22.5 R
 P a g e | 313

2995 22.5 R

3100 24 R

n1 = 12

n2 = 12

W1 = 1 + 4 + 5 + 6 + 7 + 9 + 11 + 12 + 14 + 15 + 19.5 + 19.5 =
123

n1  n2 (12)(12)
 
2 2
= 72

n1  n2 (n1  n2  1) (12)(12)( 25)

  
12 12
= 17.32

n1 (n1  1) (12)(13)
U  n1  n2   W1  (12)(12)   123
2 2
= 99

U   99  72
z 
 17.32
= 1.56

 = .05 /2 = .025

z.025 = +1.96
 P a g e | 314

Since the observed z = 1.56 < z.025 = 1.96, the decision is to fail to
reject the

null hypothesis.
 P a g e | 315

13.11 Ho: Males do not earn more than females

Ha: Males do earn more than females

Earnings Rank Gender

$28,900 1 F

31,400 2 F

36,600 3 F

40,000 4 F

40,500 5 F

41,200 6 F

42,300 7 F

42,500 8 F

44,500 9 F

45,000 10 M

47,500 11 F

47,800 12.5 F

47,800 12.5 M

48,000 14 F

50,100 15 M

51,000 16 M

51,500 17.5 M

51,500 17.5 M

53,850 19 M

55,000 20 M

57,800 21 M
 P a g e | 316

61,100 22 M

63,900 23 M

n1 = 11 n2 = 12

W1 = 10 + 12.5 + 15 + 16 + 17.5 + 17.5 + 19 + 20 + 21 + 22 + 23 =

193.5

n1  n2 (11)(12) n1  n2 (n1  n2  1) (11)(12)( 24)

    
2 2 12 12
= 66 and =
16.25

n1 (n1  1) (11)(12)
U  n1  n2   W1  (11)(12)   193.5
2 2
= 4.5

U   4.5  66
z 
 16.25
= -3.78

 = .01, z.01 = 2.33

Since the observed z = 3.78 > z.01 = 2.33, the decision is to reject
the null

hypothesis.

13.12 H0: There is no difference in the price of a single-family home in

Denver and
 P a g e | 317

Hartford

Ha: There is a difference in the price of a single-family home in Denver

and

Hartford

Price Rank City

132,405 1 D

134,127 2 H

134,157 3 D

134,514 4 H

135,062 5 D

135,238 6 H

135,940 7 D
136,333 8 H

136,419 9 H

136,981 10 D

137,016 11 D

137,359 12 H

137,741 13 H

137,867 14 H

138,057 15 D

139,114 16 H

139,638 17 D

140,031 18 H

140,102 19 D

140,479 20 D
 P a g e | 318

141,408 21 D

141,730 22 D

141,861 23 D

142,012 24 H

142,136 25 H

143,947 26 H

143,968 27 H

144,500 28 H

n1 = 13

n2 = 15

W1 = 1 + 3 + 5 + 7 + 10 + 11 + 15 + 17 + 19 +

20 + 21 + 22 + 23 = 174

n1 ( n1  1) (13)(14)
U  n1  n2   W1  (13)(15)   174
2 2
= 112

n1  n2 (13)(15)
 
2 2
= 97.5

n1  n2 (n1  n2  1) (13)(15)( 29)

  
12 12
= 21.708
 P a g e | 319

U   112  97.5
z 
 21.708
= 0.67

For  = .05 and a two-tailed test, /2 = .025 and

z.025 = + 1.96.

Since the observed z = 0.67 < z.025 = 1.96, the decision is to fail
to reject

the null hypothesis. There is not enough evidence to declare

that there is

a price difference for single family homes in Denver and

Hartford.

13.13 Ho: The population differences = 0

Ha: The population differences  0

1 2 d Rank

212 179 33 15

234 184 50 16

219 213 6 7.5

199 167 32 13.5

194 189 5 6

206 200 6 7.5

 P a g e | 320

234 212 22 11

225 221 4 5

220 223 -3 - 3.5

218 217 1 1

234 208 26 12

212 215 -3 -3.5

219 187 32 13.5

196 198 -2 -2

178 189 -11 -9

213 201 12 10

n = 16

T- = 3.5 + 3.5 + 2 + 9 = 18

(n)( n  1) (16)(17)
 
4 4
= 68

n(n  1)( 2n  1) 16(17)(33)

  
24 24
= 19.34

T   18  68
z 
 19.34
= -2.59

 = .10 /2 = .05 z.05 = ±1.645

 P a g e | 321

Since the observed z = -2.59 < z.05 = -1.645, the decision is to reject
the null

hypothesis.

13.14 Ho: Md = 0

Ha: Md  0

Before After d Rank

49 43 6 +9

41 29 12 +12

47 30 17 +14

39 38 1 + 1.5

53 40 13 +13

51 43 8 +10

51 46 5 + 7.5

49 40 9 +11

38 42 -4 - 5.5

54 50 4 + 5.5

46 47 -1 - 1.5

50 47 3 +4

44 39 5 + 7.5

49 49 0

45 47 -2 -3
 P a g e | 322

n = 15 but after dropping the zero difference, n = 14

 = .05, for two-tailed /2 = .025, and from Table A.14, T.025,14= 21

T+ = 9 + 12+ 14 + 1.5 + 13 + 10 + 7.5 + 11 + 5.5 + 4 + 7.5 = 95

T- = 5.5 + 1.5 + 3 = 10

T = min(T+,T-) = min(95, 10) = 10

Since the observed value of T = 10 < T.025, 14 = 21, the decision is to

reject the

null hypothesis. There is a significant difference in before and after.

 P a g e | 323

13.15 Ho: The population differences > 0

Ha: The population differences < 0

Before After d Rank

10,500 12,600 -2,100 -11

8,870 10,660 -1,790 -9

12,300 11,890 410 3

10,510 14,630 -4,120 -17

5,570 8,580 -3,010 -15

9,150 10,115 -965 -7

11,980 14,320 -2,370 -12

6,740 6,900 -160 -2

7,340 8,890 -1,550 -8

13,400 16,540 -3,140 -16

12,200 11,300 900 6

10,570 13,330 -2,760 -13

9,880 9,990 -110 -1

12,100 14,050 -1,950 -10

9,000 9,500 -500 -4

11,800 12,450 -650 -5

10,500 13,450 -2,950 -14

Since n = 17, use the large sample test

T+ = 3 + 6 = 9
 P a g e | 324

T =9

(n)( n  1) (17)(18)
 
4 4
= 76.5

n(n  1)( 2n  1) 17(18)(35)

  
24 24
= 21.12

T   9  76.5
z 
 21.12
= -3.20

 = .05 z.05 = -1.645

Since the observed z = -3.20 < z.05 = -1.645, the decision is to reject
the null

hypothesis.

13.16 Ho:Md = 0

Ha:Md < 0

Manual Scanner d Rank

426 473 -47 -11

 P a g e | 325

387 446 -59 -13

410 421 -11 -5.5

506 510 -4 -2

411 465 -54 -12

398 409 -11 -5.5

427 414 13 7

449 459 -10 -4

407 502 -95 -14

438 439 -1 -1

418 456 -38 -10

482 499 -17 -8

512 517 -5 -3

402 437 -35 -9

n = 14

T+ = (+7)

T- = (11 + 13 + 5.5 + 3 + 12 + 5.5 + 4 + 14 + 1 + 10 + 8 + 3 + 9)=

98

T = min(T+,T-) = min(7, 98) = 7

from Table A.14 with  = .05, n = 14, T.05,14 = 26

Since the observed T = 7 < T.05,14 = 26, the decision is to reject the null

hypothesis.

The differences are significantly less than zero and the after scores are
 P a g e | 326

significantly higher.
 P a g e | 327

13.17 Ho: The population differences 0

Ha: The population differences < 0

1999 2006 d Rank

49 54 -5 -7.5

27 38 -11 -15

39 38 1 2

75 80 -5 -7.5

59 53 6 11

67 68 -1 -2

22 43 -21 -20

61 67 -6 -11

58 73 -15 -18

60 55 5 7.5

72 58 14 16.5

62 57 5 7.5

49 63 -14 -16.5

48 49 -1 -2

19 39 -20 -19

32 34 -2 -4.5

60 66 -6 -11

80 90 -10 -13.5

55 57 -2 -4.5

68 58 10 13.5
 P a g e | 328

n = 20

T+ = 2 + 11 + 7.5 + 16.5 + 7.5 + 13.5 = 58

T = 58

(n)( n  1) (20)( 21)

 
4 4
= 105

n(n  1)( 2n  1) 20(21)( 41)

  
24 24
= 26.79

T   58  105
z 
 26.79
= -1.75

For  = .10, z.10 = -1.28

Since the observed z = -1.75 < z.10 = -1.28, the decision is to reject
the null

hypothesis.

13.18 Ho: The population differences < 0

Ha: The population differences > 0

April April
 P a g e | 329

2002 2006 d Rank

63.1 57.1 5.7 16

67.1 66.4 0.7 3.5

65.5 61.8 3.7 12

68.0 65.3 2.7 8.5

66.6 63.5 3.1 10.5

65.7 66.4 -0.7 -3.5

69.2 64.9 4.3 14

67.0 65.2 1.8 6.5

65.2 65.1 0.1 1.5

60.7 62.2 -1.5 -5

63.4 60.3 3.1 10.5

59.2 57.4 1.8 6.5

62.9 58.2 4.7 15

69.4 65.3 4.1 13

67.3 67.2 0.1 1.5

66.8 64.1 2.7 8.5

n = 16

T- = 8.5

T = 8.5

(n)( n  1) (16)(17)
 
4 4
= 68
 P a g e | 330

n(n  1)( 2n  1) 16(17)(33)

  
24 24
= 19.339

T   8.5  68
z 
 19.339
= -3.08

For  = .05, z.05 = 1.645

 3.08
Since the observed z = > z.05 = 1.645, the decision is to reject
the null

hypothesis.
 P a g e | 331

13.19 Ho: The 5 populations are identical

Ha: At least one of the 5 populations is different

1 2 3 4 5__

157 165 219 286 197

188 197 257 243 215

175 204 243 259 235

174 214 231 250 217

201 183 217 279 240

203 203 233

213

BY RANKS

1 2 3 4 5__

1 2 18 29 7.5

6 7.5 26 23.5 15

4 12 23.5 27 21

3 14 19 25 16.5

9 5 16.5 28 22

10.5 10.5 20

_ _ __ __ 13_

Tj 33.5 40.5 113.5 132.5 115

nj 6 5 6 5 7
 P a g e | 332
2
Tj (33.5) 2 (40.5) 2 (113.5) 2 (132.5) (115) 2
n 
6

5

6

5

7
j

= 8,062.67

n = 29

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
29(30)
(8,062.67)  3(30)

= 21.21

 = .01 df = c - 1 = 5 - 1 = 4

2.01,4 = 13.2767

Since the observed K = 21.21 > 2.01,4 = 13.2767, the decision is to

reject the null

hypothesis.
 P a g e | 333

13.20 Ho: The 3 populations are identical

Ha: At least one of the 3 populations is different

Group 1 Group 2 Group 4

19 30 39

21 38 32

29 35 41

22 24 44

37 29 30

42 27

33

By Ranks

Group 1 Group 2 Group 3

1 8.5 15

2 14 10

6.5 12 16

3 4 18

13 6.5 8.5

17 5

__ _ 11_

Tj 42.5 45 83.5

nj 6 5 7
 P a g e | 334

2
Tj (42.5) 2 (45) 2 (83.5) 2
n 
6

5

7
j

= 1,702.08

n = 18

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
18(19)
(1,702.08)  3(19)

= 2.72

 = .05, df = c - 1 = 3 - 1 = 2

2.05,2 = 5.9915

Since the observed K = 2.72 < 2.05,2 = 5.9915, the decision is to fail
to reject

the null hypothesis.

 P a g e | 335

13.21 Ho: The 4 populations are identical

Ha: At least one of the 4 populations is different

Region 1 Region 2 Region 3 Region 4

$1,200 $225 $ 675 $1,075

450 950 500 1,050

110 100 1,100 750

800 350 310 180

375 275 660 330

200 680

425

By Ranks

Region 1 Region 2 Region 3 Region 4

23 5 15 21

12 19 13 20

2 1 22 17

18 9 7 3

10 6 14 8

4 16

_ _ _ 11

T j 69 40 71 96

nj 6 5 5 7
 P a g e | 336

2
Tj (69) 2 (40) 2 (71) 2 (96) 2
n 
6

5

5

7
j

= 3,438.27

n = 23

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
23(24)
(3,428.27)  3(24)

= 2.75

 = .05 df = c - 1 = 4 - 1 = 3

2.05,3 = 7.8147

Since the observed K = 2.75 < 2.05,3 = 7.8147, the decision is to fail
to reject

the null hypothesis.

13.22 Ho: The 3 populations are identical

Ha: At least one of the 3 populations is different

Small Town City Suburb

$21,800 $22,300 $22,000

 P a g e | 337

22,500 21,900 22,600

21,750 21,900 22,800

22,200 22,650 22,050

21,600 21,800 21,250

22,550

By Ranks

Small Town City Suburb

4.5 11 8

12 6.5 14

3 6.5 16

10 15 9

2 4.5 1

__ __ 13

Tj 31.5 43.5 61

nj 5 5 6

2
Tj (31.5) 2 (43.5) 2 (61) 2
 nj

5

5

6
= 1,197.07

n = 16
 P a g e | 338
2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
16(17)
(1,197.07)  3(17)

= 1.81

 = .05 df = c - 1 = 3 - 1 = 2

2.05,2 = 5.9915

Since the observed K = 1.81 < 2.05,2 = 5.9915, the decision is to fail
to reject

the null hypothesis.

 P a g e | 339

13.23 Ho: The 4 populations are identical

Ha: At least one of the 4 populations is different

Amusement Parks Lake Area City National

Park

0 3 2 2

1 2 2 4

1 3 3 3

0 5 2 4

2 4 3 3

1 4 2 5

0 3 3 4

5 3 4

2 1

By Ranks

Amusement Parks Lake Area City

National Park

2 20.5 11.5 11.5

5.5 11.5 11.5 28.5

5.5 20.5 20.5 20.5

2 33 11.5 28.5

11.5 28.5 20.5 20.5

5.5 28.5 11.5 33

2 20.5 20.5 28.5

 P a g e | 340

33 20.5 28.5

11.5 5.5

__ __ 20.5 ____

Tj 34 207.5 154.0
199.5

nj 7 9 10 8

2
Tj (34) 2 (207.5) 2 (154) 2 (199.5)
n 
7

9

10

8
j

= 12,295.80

n = 34

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
34(35)
(12,295.80)  3(35)

= 18.99

 = .05 df = c - 1 = 4 - 1 = 3

2.05,3 = 7.8147

Since the observed K = 18.99 > 2.05,3 = 7.8147, the decision is to

reject the

null hypothesis.

13.24 Ho: The 3 populations are identical

Ha: At least one of the 3 populations is different

 P a g e | 341

Day Shift Swing Shift Graveyard Shift

52 45 41

57 48 46

53 44 39

56 51 49

55 48 42

50 54 35

51 49 52

43

By Ranks

Day Shift Swing Shift Graveyard Shift

16.5 7 3

22 9.5 8

18 6 2

21 14.5 11.5

20 9.5 4

13 19 1

14.5 11.5 16.5

_ 5 _ ___

Tj 125 82 46

nj 7 8 7

2
Tj (125) 2 (82) 2 (46) 2
n 
7

8

7
j

= 3,374.93
 P a g e | 342

n = 22

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
22(23)
(3,374.93)  3(23)

= 11.04

 = .05 df = c - 1 = 3 - 1 = 2

2.05,2 = 5.9915

Since the observed K = 11.04 > 2.05,2 = 5.9915, the decision is to

reject the

null hypothesis.
 P a g e | 343

13.25 Ho: The treatment populations are equal

Ha: At least one of the treatment populations yields larger values than
at least one

other treatment population.

Use the Friedman test with  = .05

c = 5, b = 5, df = c - 1 = 4, 2.05,4 = 9.4877

If the observed value of 2 > 9.4877, then the decision will be to reject
the null

hypothesis.

Shown below are the data ranked by blocks:

1 2 3 4 5

1 1 4 3 5 2

2 1 3 4 5 2

3 2.5 1 4 5 2.5

4 3 2 4 5 1

5 4 2 3 5 1

Rj 11.5 12 18 25
8.5
 P a g e | 344

Rj2 132.25 144 324 625

72.25

Rj2 = 1,297.5

12 12

2
r 
2
R j  3b(c  1)  (1,297.5)  3(5)(6)
bc(c  1) (5)(5)(6)
= 13.8

Since the observed value of r2 = 13.8 > 4,.052 = 9.4877, the decision
is to

reject the null hypothesis. At least one treatment population yields

larger values than at least one other treatment population.
 P a g e | 345

13.26 Ho: The treatment populations are equal

Ha: At least one of the treatment populations yields larger values than
at least one

other treatment population.

Use the Friedman test with  = .05

c = 6, b = 9, df = c - 1 = 5, 2.05,5 = 11.0705

If the observed value of 2 > 11.0705, then the decision will be to

reject the null hypothesis.

Shown below are the data ranked by blocks:

1 2 3 4 5 6

1 1 3 2 6 5 4

2 3 5 1 6 4 2

3 1 3 2 6 5 4

4 1 3 4 6 5 2

5 3 1 2 4 6 5

6 1 3 2 6 5 4

7 1 2 4 6 5 3

8 3 1 2 6 5 4

9 1 2 3 6 5 4

Rj 15 25 25 56 50 38
 P a g e | 346

Rj2 225 625 625 3136 2500 1444

Rj2 = 8,555.5

12 12

2
r 
2
R j  3b(c  1)  (8,555)  3(9)(7)
bc(c  1) (9)(6)(7)
= 82.59

Since the observed value of r2 = 82.59 > 5,.052 = 11.0705, the
decision is to

reject the null hypothesis. At least one treatment population yields

larger values

than at least one other treatment population.

 P a g e | 347

13.27 Ho: The treatment populations are equal

Ha: At least one of the treatment populations yields larger values

than at least one

other treatment population.

Use the Friedman test with  = .01

c = 4, b = 6, df = c - 1 = 3, 2.01,3 = 11.3449

If the observed value of 2 > 11.3449, then the decision will be to

reject the null hypothesis.

Shown below are the data ranked by blocks:

1 2 3 4

1 1 4 3 2

2 2 3 4 1

3 1 4 3 2

4 1 3 4 2

5 1 3 4 2

6 2 3 4 1

Rj 8 20 22 10

Rj2 64 400 484 100

 P a g e | 348

Rj2 = 1,048

12 12

2
r2  R j  3b(c  1)  (1,048)  3(6)(5)
bc(c  1) (6)( 4)(5)
= 14.8

Since the observed value of r2 = 14.8 > 23,.01 = 11.3449, the decision
is to

reject the null hypothesis. At least one treatment population yields

larger values

than at least one other treatment population.

 P a g e | 349

13.28 Ho: The treatment populations are equal

Ha: At least one of the treatment populations yields larger values than
at least one

other treatment population.

Use the Friedman test with  = .05

c = 3, b = 10, df = c - 1 = 2, 2.05,2 = 5.9915

If the observed value of 2 > 5.9915, then the decision will be to reject
the null

hypothesis.

Shown below are the data ranked by blocks:

Worker 5-day 4-day 3.5 day

1 3 2 1

2 3 2 1

3 3
1
2
4 3 2 1

5 2 3 1

6 3 2 1

7 3 1 2

8 3 2 1

9 3 2 1
 P a g e | 350

10 3 1 2

Rj 29 18 13

Rj2 841 324 169

Rj2 = 1,334

12 12

2
r2  R j  3b(c  1)  (1,334)  3(10)( 4)
bc (c  1) (10)(3)( 4)
= 13.4

Since the observed value of r2 = 13.4 > 2.05,2 = 5.9915, the decision
is to

reject the null hypothesis. At least one treatment population yields

larger values

than at least one other treatment population.

 P a g e | 351

13.29 c = 4 treatments b = 5 blocks

S = r2 = 2.04 with a p-value of .564.

Since the p-value of .564 >  = .10, .05, or .01, the decision is to fail
to reject

the null hypothesis. There is no significant difference in treatments.

13.30 The experimental design is a random block design that has been
analyzed using a Friedman test. There are five treatment levels and
seven blocks. Thus, the degrees of freedom are four. The observed
value of S = 13.71 is the equivalent of r2. The p-value is .009
indicating that this test is significant at alpha .01. The null hypothesis
is rejected. That is, at least one population yields larger values than at
least one other population. An examination of estimated medians
shows that treatment 1 has the lowest value and treatment 3 has the
highest value.

13.31 x y x Ranked y Ranked d d2

23 201 3 2 1 1

41 259 10.5 11 -.5 0.25

37 234 8 7 1 1

29 240 6 8 -2 4
 P a g e | 352

25 231 4 6 -2 4

17 209 1 3 -2 4

33 229 7 5 2 4

41 246 10.5 9 1.5 2.25

40 248 9 10 -1 1

28 227 5 4 1 1

19 200 2 1 1 1

d2 = 23.5

n = 11

6 d 2 6(23.5)
rs  1   1
n(n  1)
2
11(120)
= .893
 P a g e | 353

13.32 x y d d2

4 6 -2 4

5 8 -3 9

8 7 1 1

11 10 1 1

10 9 1 1

7 5 2 4

3 2 1 1

1 3 -2 4

2 1 1 1

9 11 -2 4

6 4 2 4

d2 = 34

n = 11

6 d 2 6(34)
rs  1   1
n(n  1)
2
11(120)
= .845

13.33 x y x Ranked y Ranked d d2

 P a g e | 354

99 108 8 2 6 36

67 139 4 5 -1 1

82 117 6 3 3 9

46 168 1 8 -7 49

80 124 5 4 1 1

57 162 3 7 -4 16

49 145 2 6 -4 16

91 102 7 1 6 36

d2 = 164

n =8

6 d 2 6(164)
rs  1   1
n(n  1)
2
8(63)
= -.95
 P a g e | 355

13.34 x y x Ranked y Ranked d d2

92 9.3 8 9 -1 1

96 9.0 9 8 1 1

91 8.5 6.5 7 -.5 .25

89 8.0 5 3 2 4

91 8.3 6.5 5 1.5 2.25

88 8.4 4 6 -2 4

84 8.1 3 4 -1 1

81 7.9 1 2 -1 1

83 7.2 2 1 1 1

d2 = 15.5

n =9

6 d 2 6(15.5)
rs  1   1
n(n  1)
2
9(80)
= .871

13.35 Bank Home Bank Home

Credit Equity Cr. Cd. Eq. Loan

Card Loan Rank Rank d d2

2.51 2.07 12 1 11 121

2.86 1.95 6.5 2 4.5 20.25

2.33 1.66 13 6 7 49
 P a g e | 356

2.54 1.77 10 3 7 49

2.54 1.51 10 7.5 2.5 6.25

2.18 1.47 14 10 4 16

3.34 1.75 3 4 -1 1

2.86 1.73 6.5 5 1.5 2.25

2.74 1.48 8 9 -1 1

2.54 1.51 10 7.5 2.5 6.25

3.18 1.25 4 14 -10 100

3.53 1.44 1 11 -10 100

3.51 1.38 2 12 -10 100

3.11 1.30 5 13 -8 64

d2 = 636

n = 14

6 d 2 6(636)
rs  1   1
n(n  1)
2
14(142  1)
= -.398

There is a very modest negative correlation between overdue

payments for bank

credit cards and home equity loans.

13.36 Iron Steel

Year Rank Rank d d2

1 12 12 0 0
 P a g e | 357

2 11 10 1 1

3 3 5 -2 4

4 2 7 -5 25

5 4 6 -2 4

6 10 11 -1 1

7 9 9 0 0

8 8 8 0 0

9 7 4 3 9

10 1 3 -2 4

11 6 2 4 16

12 5 1 4 16

d2 = 80

6 d 2 6(80)
rs  1   1
n(n  1)
2
12(144  1)
= 0.72

13.37 No. Co. No. Eq. Is.

on NYSE on AMEX Rank NYSE Rank AMEX d d2

1774 1063 11 1 10 100

1885 1055 10 2 8 64

2088 943 95 4 16
2361 1005 8 3 5 25
 P a g e | 358

2570 981 7 4 3 9

2675 936 6 6 0 0

2907 896 4 7 -3 9

3047 893 2 8 -6 36

3114 862 19 -8 64
3025 769 3 10 -7 49
2862 765 5 11 -6
36 d2 = 162

d 2

= 408

n = 11

6 d 2 6( 408)
rs  1   1
n(n  1)
2
11(112  1)
= -0.855

There is a strong negative correlation between the number of

companies listed

on the NYSE and the number of equity issues on the American Stock
Exchange.

13.38  = .05

H 0: The observations in the sample are randomly generated

Ha: The observations in the sample are not randomly generated

 P a g e | 359

n1 = 13, n2 = 21

R = 10

Since this is a two-tailed test, use /2 = .025. The critical value is: z.025
= + 1.96

2n1n2 2(13)( 21)

R  1  1
n1  n2 13  21
= 17.06

2n1n2 (2n1n2  n1  n2 ) 2(13)( 21) 2(13)( 21)  13  21

R  
(n1  n2 ) 2 (n1  n2  1) (13  21) 2 (13  21  1)
= 2.707

R   R 10  17.06
z 
R 2.707
= -2.61

Since the observed z = - 2.61 < z.025 = - 1.96, the decision is to reject
the null

hypothesis. The observations in the sample are not randomly

generated.
 P a g e | 360

13.39 Sample 1 Sample 2

573 547

532 566

544 551

565 538

540 557

548 560

536 557

523 547

 = .01 Since n1 = 8, n2 = 8 < 10, use the small sample Mann-

Whitney U test.
 P a g e | 361

x Rank Group

523 1 1

532 2 1

536
3 1
538 4 2

540 5 1

544 6 1

547 7.5 2

547 7.5 2

548 9 1

551 10 2

557 11.5 2

557 11.5 2

560 13 2

565 14 1

566 15 2

573 16 1

W1 = 1 + 2 + 3 + 5 + 6 + 9 + 14 + 16 = 56

n1 (n1  1) (8)(9)
U1  n1  n2   W1  (8)(8)   56
2 2
= 44
 P a g e | 362

U 2  n1  n2  U1
= 8(8) - 44 = 20

Take the smaller value of U1, U2 = 20

From Table A.13, the p-value (1-tailed) is .1172, for 2-tailed, the p-value
is .2344. Since the p-value is >  = .05, the decision is to fail to reject
the null hypothesis.
 P a g e | 363

13.40  = .05, n = 9

H 0: Md = 0

Ha: Md  0

Group 1 Group 2 d Rank

5.6 6.4 -0.8 -8.5

1.3 1.5 -0.2 -4.0

4.7 4.6 0.1 2.0

3.8 4.3 -0.5 -6.5

2.4 2.1 0.3 5.0

5.5 6.0 -0.5 -6.5

5.1 5.2 -0.1 -2.0

4.6 4.5 0.1 2.0

3.7 4.5 -0.8 -8.5

Since n = 9, from Table A.14 (2-tailed test), T.025 = 6

T+ = 2 + 5 + 2 = 9

T- = 8.5 + 4 + 6.5 + 6.5 + 2 + 8.5 = 36

T = min(T+, T-) = 9
 P a g e | 364

Since the observed value of T = 9 > T.025 = 6, the decision is to fail to

reject the

null hypothesis. There is not enough evidence to declare that there

is a difference between the two groups.
 P a g e | 365

13.41 nj = 7, n = 28, c = 4, df = 3

Group 1 Group 2 Group 3

Group 4

6 4 3 1

11 13 7 4

8 6 7 5

10 8 5 6

13 12 10 9

7 9 8 6

10 8 5 7

By Ranks:

Group 1 Group 2 Group 3 Group 4

9.5 3.5 2 1

25 27.5 13.5 3.5

17.5 9.5 13.5 6

23 17.5 6 9.5

27.5 26 23 20.5

13.5 20.5 17.5 9.5

23 17.5 6 13.5

Tj 139 122 81.5 63.5

 P a g e | 366
2
Tj
n
j

= 2760.14 + 2126.29 + 948.89 + 576.04 = 6411.36

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
28(29)
(6411.36)  3(29)

= 7.75

The critical value of chi-square is: 23,.01 = 11.3449.

Since K = 7.75 < 23,.01 = 11.3449, the decision is to fail to reject the
null

hypothesis.
 P a g e | 367

13.42  = .05, b = 7, c = 4, df = 3

2.05,3 = 7.8147

H0: The treatment populations are equal

Ha: At least one treatment population yields larger values than at least
one other

treatment population

Blocks Group 1 Group 2 Group 3 Group 4

1 16 14 15 17

2 8 6 5 9

3 19 17 13 9

4 24 26 25 21

5 13 10 9 11

6 19 11 18 13

7 21 16
14 15

By Ranks:

Blocks Group 1 Group 2 Group 3

Group 4

1 3 1 2 4

2 3 2 1 4

3 4 3 2 1

4 2 4 3 1
 P a g e | 368

5 4 2 1 3

6 4 1 3 2

7 4 3 1 2

Rj 24 16 13 17

Rj2 576 256 169 289

Rj2 = 567 + 256 + 169 + 289 = 1290

12 12

2
r 
2
R j  3b(C  1)  (1,290)  3(7)(5)
bC (C  1) (7)( 4)(5)
= 5.57

Since r2 = 5.57 < 2.05,3 = 7.8147, the decision is to fail to reject
the null

hypothesis. The treatment population means are equal.

 P a g e | 369

13.43 Ranks

1 2 1 2 d d2

101 87 1 7 -6 36

129 89 2 8 -6 36

133 84 3 6 -3 9

147 79 4 5 -1 1

156 70 5 3 2 4

179 64 6 1 5 25

183 67 7 2 5 25

190 71 8 4 4 16

d2 = 152

n=8

6 d 2 6(152)
rs  1   1
n(n  1)
2
8(63)
= -.81

13.44 Ho: The 3 populations are identical

Ha: At least one of the 3 populations is different

1 Gal. 5 Gal. 10 Gal.

1.1 2.9 3.1

 P a g e | 370

1.4 2.5 2.4

1.7 2.6 3.0

1.3 2.2 2.3

1.9 2.1 2.9

1.4 2.0 1.9

2.1 2.7

By Ranks

1 Gal. 5 Gal. 10 Gal.

1 17.5 20

3.5 14 13

5 15 19

2 11 12

6.5 9.5 17.5

3.5 8 6.5

9.5 16

Tj 31 91 88

nj 7 7 6

2
Tj (31) 2 (91) 2 (88) 2
n 
7

7

6
j

= 2,610.95

n = 20
 P a g e | 371

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
20(21)
(2,610.95)  3(21)

= 11.60

 = .01 df = c - 1 = 3 - 1 = 2

2.01,2 = 9.2104

Since the observed K = 11.60 > 2.01,2 = 9.2104, the decision is to

reject the null

hypothesis.

13.45 N = 40 n1 = 24 n2 = 16  = .05

Use the large sample runs test since both n1, n2 are not less than 20.

H0: The observations are randomly generated

Ha: The observations are not randomly generated

With a two-tailed test, /2 = .025, z.025 = + 1.96. If the observed z > .
196

or < -1.96, the decision will be to reject the null hypothesis.

 P a g e | 372

R = 19

2n1n2 2(24)(16)
R  1  1
n1  n2 24  16
= 20.2

2n1n2 (2n1n2  n1  n2 ) 2(24)(16) 2(24)(16)  24  16

R  
(n1  n2 ) 2 (n1  n2  1) (40) 2 (39)
= 2.993

R   R 19  20.2
z 
R 2.993
= -0.40

Since z = -0.40 > z.025 = -1.96, the decision is to fail to reject the
null hypothesis.

13.46 Use the Friedman test. Let  = .05

H0: The treatment populations are equal

Ha: The treatment populations are not equal

c = 3 and b = 7

Operator Machine 1 Machine 2 Machine 3

1 231 229 234

 P a g e | 373

2 233 232 231

3 229 233 230

4 232 235 231

5 235 228 232

6 234 237 231

7 236 233 230

By ranks:

Operator Machine 1 Machine 2 Machine 3

1 2 1 1

3 1 3 2

4 2 3 1

5 3 1 2

6 2 3 1

7 3 2 1

Rj 16 15 11

Rj2 256 225 121

df = c - 1 = 2 2.05,2 = 5.99147.

If the observed 2r > 5.99147, the decision will be to reject the null
hypothesis.
 P a g e | 374

Rj2 = 256 + 225 + 121 = 602

12 12

2
r2  R j  3b(c  1)  (602)  3(7)( 4)
bc(c  1) (7)(3)( 4)
= 2

Since 2r = 2 < 2.05,2 = 5.99147, the decision is to fail to reject the
null

hypothesis.
 P a g e | 375

13.47 Ho: EMS workers are not older

Ha: EMS workers are older

Age Rank Group

21 1 1

23 2 1

24 3 1

25 4 1

27 6 1

27 6 2

27 6 2

28 9 1

28 9 2

28 9 2

29 11 2

30 13 2

30 13 2

30 13 2

32 15 1

33 16.5 2

33 16.5 2

36 18.5 1

36 18.5 2

37 20 1

39 21 2
 P a g e | 376

41 22 1

n1 = 10 n2 = 12

W1 = 1 + 2 + 3 + 4 + 6 + 9 + 15 + 18.5 + 20 + 22 = 100.5

n1  n2 (10)(12)
 
2 2
= 60

n1  n2 (n1  n2  1) (10)(12)( 23)

  
12 12
= 15.17

n1 (n1  1) (10)(11)
U  n1  n2   W1  (10)(12)   100.5
2 2
= 74.5

U   74.5  60
z 
 15.17
= 0.96 with  = .05, z.05 = -1.645

 1.645
Since the observed z = 0.96 < z.05 = , the decision is to fail to
reject the

null hypothesis.

13.48 Ho: The population differences = 0

 P a g e | 377

Ha: The population differences  0

With Without d Rank

1180 1209 -29 -6

874 902 -28 -5

1071 862 209 18

668 503 165 15

889 974 -85 -12.5

724 675 49 9

880 821 59 10

482 567 -85 -12.5

796 602 194 16

1207 1097 110 14

968 962 6 1

1027 1045 -18 -4

1158 896 262 20

670 708 -38 -8

849 642 207 17

559 327 232 19

449 483 -34 -7

992 978 14 3

1046 973 73 11

852 841 11 2

n = 20
 P a g e | 378

T- = 6 + 5 + 12.5 + 12.5 + 4 + 8 + 7 = 55

T = 55

(n)( n  1) (20)( 21)

 
4 4
= 105

n(n  1)( 2n  1) 20(21)( 41)

  
24 24
= 26.79

T   55  105
z 
 26.79
= -1.87

 = .01, /2 = .005 z.005 = ±2.575

Since the observed z = -1.87 > z.005 = -2.575, the decision is to fail to
reject the

null hypothesis.

13.49 H0: There is no difference between March and June

Ha: There is a difference between March and June

GMAT Rank Month

350 1 J
 P a g e | 379

430 2 M

460 3 J

470 4 J

490 5 M

500 6 M

510 7 M

520 8 J

530 9.5 M

530 9.5 J

540 11 M

550 12.5 M

550 12.5 J

560 14 M

570 15.5 M

570 15.5 J

590 17 J

600 18 M

610 19 J

630 20 J

n1 = 10 n2 = 10

W1 = 1 + 3 + 4 + 8 + 9.5 + 12.5 + 15.5 + 17 + 19 + 20 = 109.5

 P a g e | 380

n1 (n1  1) (10)(11)
U1  n1  n2   W1  (10)(10)   109.5
2 2
= 45.5

U 2  n1  n2  U1
= (10)(10) - 45.5 = 54.5

From Table A.13, the p-value for U = 45 is .3980 and for 44 is .3697.
For a

two-tailed test, double the p-value to at least .739. Using  = .10, the
decision is

to fail to reject the null hypothesis.

 P a g e | 381

13.50 Use the Friedman test. b = 6, c = 4, df = 3,  = .05

H0: The treatment populations are equal

Ha: At least one treatment population yields larger values than at

least on other

treatment population

The critical value is: 2.05,3 = 7.8147

Location

Brand 1 2 3 4
A 176 58 111 120

B 156 62 98 117

C 203 89 117 105

D 183 73 118 113

E 147 46 101 114

F 190 83 113 115

By ranks:

Location

Brand 1 2 3 4

A 4 1 2 3

B 4 1 2 3

C 4 1 3 2

D 4 1 3 2
 P a g e | 382

E 4 1 2 3

F 4 1 2 3

Rj 24 6 14 16

Rj2 576 36 196 256

Rj2 = 1,064

12 12

2
r2  R j  3b(c  1)  (1,064)  3(6)(5)
bc(c  1) (6)( 4)(5)
= 16.4

Since r2 = 16.4 > 2.05,3 = 7.8147, the decision is to reject the null
hypothesis.

At least one treatment population yields larger values than at least one
other

treatment population. An examination of the data shows that location

one produced the highest sales for all brands and location two
produced the lowest sales of gum for all brands.
 P a g e | 383

13.51 Ho: The population differences = 0

Ha: The population differences  0

Box No Box d Rank

185 170 15 11

109 112 -3 -3

92 90 2 2

105 87 18 13.5

60 51 9 7

45 49 -4 -4.5

25 11 14 10

58 40 18 13.5

161 165 -4 -4.5

108 82 26 15.5

89 94 -5 -6

123 139 -16 -12

34 21 13 8.5

68 55 13 8.5

59 60 -1 -1

78 52 26 15.5

n = 16

T- = 3 + 4.5 + 4.5 + 6 + 12 + 1 = 31

T = 31
 P a g e | 384

(n)( n  1) (16)(17)
 
4 4
= 68

n(n  1)( 2n  1) 16(17)(33)

  
24 24
= 19.34

T   31  68
z 
 19.34
= -1.91

 = .05, /2 = .025 z.025 = ±1.96

Since the observed z = -1.91 > z.025 = -1.96, the decision is to fail to
reject the

null hypothesis.
 P a g e | 385

13.52 Ranked Ranked

Cups Stress Cups Stress d d2

25 80 6 8 -2 4

41 85 9 9 0 0

16 35 4 3 1 1

0 45 1 5 -4 16

11 30 3 2 1 1

28 50 7 6 1 1

34 65 8 7 1 1

18 40 5 4 1 1

5 20 2 1 1 1

d2 = 26

n=9

6 d 2 6(26)
rs  1   1
n(n  1)
2
9(80)
= .783

13.53 n1 = 15, n2 = 15 Use the small sample Runs test

 = .05, /.025
 P a g e | 386

H 0: The observations in the sample were randomly generated.

Ha: The observations in the sample were not randomly generated

From Table A.11, lower tail critical value = 10

From Table A.12, upper tail critical value = 22

R = 21

Since R = 21 between the two critical values, the decision is to fail to

reject the

null hypothesis. The observations were randomly generated.

 P a g e | 387

13.54 Ho: The population differences > 0

Ha: The population differences < 0

Before After d Rank

430 465 -35 -11

485 475 10 5.5

520 535 -15 - 8.5

360 410 -50 -12

440 425 15 8.5

500 505 -5 -2

425 450 -25 -10

470 480 -10 -5.5

515 520 -5 -2

430 430 0 OMIT

450 460 -10 -5.5

495 500 -5 -2

540 530 10 5.5

n = 12

T+ = 5.5 + 8.5 + 5.5 = 19.5 T = 19.5

From Table A.14, using n = 12, the critical T for  = .01, one-tailed, is
10.
 P a g e | 388

Since T = 19.5 is not less than or equal to the critical T = 10, the
decision is to fail

to reject the null hypothesis.

 P a g e | 389

13.55 Ho: With ties have no higher scores

Ha: With ties have higher scores

Rating Rank Group

16 1 2

17 2 2

19 3.5 2

19 3.5 2

20 5 2

21 6.5 2

21 6.5 1

22 9 1

22 9 1

22 9 2

23 11.5 1

23 11.5 2

24 13 2

25 15.5 1

25 15.5 1

25 15.5 1

25 15.5 2

26 19 1

26 19 1

26 19 2

27 21 1

28 22 1
 P a g e | 390

n1 = 11 n2 = 11

W1 = 6.5 + 9 + 9 + 11.5 + 15.5 + 15.5 + 15.5 + 19 + 19 + 21 + 22 =

163.5

n1  n2 (11)(11)
 
2 2
= 60.5

n1  n2 (n1  n2  1) (11)(11)( 23)

  
12 12
= 15.23

n1 (n1  1) (11)(12)
U  n1  n2   W1  (11)(11)   163.5
2 2
= 23.5

U   23.5  60.5
z 
 15.23
= -2.43 For  = .05, z.05 = 1.645

 2.43
Since the observed z = > z.05 = 1.645, the decision is to reject
the null

hypothesis.

13.56 Ho: Automatic no more productive

Ha: Automatic more productive

 P a g e | 391

Sales Rank Type of Dispenser

92 1 M

105 2 M

106 3 M

110 4 A

114 5 M

117 6 M

118 7.5 A

118 7.5 M

125 9 M

126 10 M

128 11 A

129 12 M

137 13 A

143 14 A

144 15 A

152 16 A

153 17 A

168 18 A

n1 = 9 n2 = 9

W1 = 4 + 7.5 + 11 + 13 + 14 + 15 + 16 + 17 + 18 = 115.5
 P a g e | 392

n1 (n1  1) (9)(10)
U1  n1  n2   W1  (9)(9)   115.5
2 2
= 10.5

U 2  n1  n2  U1
= 81 – 10.5 = 70.5

The smaller of the two is U1 = 10.5

 = .01

From Table A.13, the p-value = .0039. The decision is to reject the
null

hypothesis since the p-value is less than .01.

 P a g e | 393

13.57 Ho: The 4 populations are identical

Ha: At least one of the 4 populations is different

45 55 70 85

216 228 219 218

215 224 220 216

218 225 221 217

216 222 223 221

219 226 224 218

214 225 217

By Ranks:

45 55 70 85

4 23 11.5 9

2 18.5 13 4

9 20.5 14.5 6.5

4 16 17 14.5

11.5 22 18.5 9

1 20.5 6.5

Tj 31.5 120.5 74.5 49.5

nj 6 6 5 6
 P a g e | 394
2
Tj (31.5) 2 (120.5) 2 (74.5) 2 (49.5)
n 
6

6

5

6
j

= 4,103.84

n = 23

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
23(24)
(4,103.84)  3(24)

= 17.21

 = .01 df = c - 1 = 4 - 1 = 3

2.01,3 = 11.3449

Since the observed K = 17.21 > 2.01,3 = 11.3449, the decision is to

reject the

null hypothesis.
 P a g e | 395

13.58

Ranks Ranks

Sales Miles Sales Miles d d2

150,000 1,500 1 1 0 0

210,000 2,100 2 2 0 0

285,000 3,200 3 7 -4 16

301,000 2,400 4 4 0 0

335,000 2,200 5 3 2 4

390,000 2,500 6 5 1 1

400,000 3,300 7 8 -1 1

425,000 3,100 8 6 2 4

440,000 3,600 9 9 0 0

d2 = 26

n=9

6 d 2 6(26)
rs  1   1
n(n  1)
2
9(80)
= .783
 P a g e | 396

13.59 Ho: The 3 populations are identical

Ha: At least one of the 3 populations is different

3-day Quality Mgmt. Inv.

9 27 16

11 38 21

17 25 18

10 40 28

22 31 29

15 19 20

6 35 31

By Ranks:

3-day Quality Mgmt. Inv.

2 14 6

4 20 11

7 13 8

3 21 15

12 17.5 16

5 9 10

1 19 17.5

Tj 34 113.5 83.5

nj 7 7 7
 P a g e | 397

2
Tj (34) 2 (113.5) 2 (83.5) 2
n 
7

7

7
j

= 3,001.5

n = 21

2
12 Tj 12
K
n(n  1)
 nj
 3(n  1) 
21(22)
(3,001.5)  3(22)

= 11.96

 = .10 df = c - 1 = 3 - 1 = 2

2.10,2 = 4.6052

Since the observed K = 11.96 > 2.10,2 = 4.6052, the decision is to

reject the

null hypothesis.

13.60 Ho: The population differences > 0

Ha: The population differences < 0

Husbands Wives d Rank

27 35 -8 -12

22 29 -7 -11
 P a g e | 398

28 30 -2 -6.5

19 20 -1 -2.5

28 27 1 2.5

29 31 -2 -6.5

18 22 -4 -9.5

21 19 2 6.5
25 29 -4 -9.5

18 28 -10 -13.5

20 21 -1 -2.5

24 22 2 6.5

23 33 -10 -13.5

25 38 -13 -16.5

22 34 -12 -15

16 31 -15 -18

23 36 -13 -16.5

30 31 -1 -2.5

n = 18

T+ = 2.5 + 6.5 + 6.5 = 15.5

T = 15.51

(n)( n  1) (18)(19)
 
4 4
= 85.5
 P a g e | 399

n(n  1)( 2n  1) 18(19)(37)

  
24 24
= 22.96

T   15.5  85.5
z 
 22.96
= -3.05

 = .01 z.01 = -2.33

Since the observed z = -3.05 < z.01 = -2.33, the decision is to reject
the null

hypothesis.

13.61 This problem uses a random block design, which is analyzed by the
Friedman

nonparametric test. There are 4 treatments and 10 blocks. The value

of the

observed r2 (shown as S) is 12.16 (adjusted for ties) and has an

associated

p-value of .007 that is significant at  = .01. At least one treatment

population

yields larger values than at least one other treatment population.

Examining the

treatment medians, treatment one has an estimated median of 20.125

and
 P a g e | 400

treatment two has a treatment median of 25.875. These two are the
farthest apart.

13.62 This is a Runs test for randomness. n1 = 21, n2 = 29. Because of the
size of the

n’s, this is a large sample Runs test. There are 28 runs, R = 28.

µR = 25.36 R = 3.34

28  25.36
z
3.34
= 0.79

The p-value for this statistic is .4387 for a two-tailed test. The decision
is to fail

to reject the null hypothesis at  = .05.

 P a g e | 401

13.63 A large sample Mann-Whitney U test is being computed. There are 16

observations in each group. The null hypothesis is that the two

populations are

identical. The alternate hypothesis is that the two populations are

not identical.

The value of W is 191.5. The p-value for the test is .0066. The test is
significant

at  = .01. The decision is to reject the null hypothesis. The two

populations are

not identical. An examination of medians shows that the median for

group two

(46.5) is larger than the median for group one (37.0).

13.64 A Kruskal-Wallis test has been used to analyze the data. The null
hypothesis is

that the four populations are identical; and the alternate hypothesis is
that at least one of the four populations is different. The H statistic
(same as the K statistic) is 11.28 when adjusted for ties. The p-value
for this H value is .010 which indicates that there is a significant
difference in the four groups at  = .05 and marginally so for  = .01.
An examination of the medians reveals that all group medians are the
same (35) except for group 2 that has a median of 25.50. It is likely
that it is group 2 that differs from the other groups.
 P a g e | 402

Chapter 14
Simple Regression Analysis

LEARNING OBJECTIVES

The overall objective of this chapter is to give you an understanding of

bivariate regression analysis, thereby enabling you to:

1. Compute the equation of a simple regression line from a sample of data

and interpret the slope and intercept of the equation.
2. Understand the usefulness of residual analysis in examining the fit of
the regression line to the data and in testing the assumptions
underlying regression analysis.

3. Compute a standard error of the estimate and interpret its meaning.

4. Compute a coefficient of determination and interpret it.

5. Test hypotheses about the slope of the regression model and interpret
the results.

6. Estimate values of y using the regression model.

7. Develop a linear trend line and use it to forecast.

 P a g e | 403

CHAPTER TEACHING STRATEGY

This chapter is about all aspects of simple (bivariate, linear) regression.

Early in the chapter through scatter plots, the student begins to understand
that the object of simple regression is to fit a line through the points. Fairly
soon in the process, the student learns how to solve for slope and y intercept
and develop the equation of the regression line. Most of the remaining
material on simple regression is to determine how good the fit of the line is
and if assumptions underlying the process are met.

The student begins to understand that by entering values of the independent

variable into the regression model, predicted values can be determined. The

question then becomes: Are the predicted values good estimates of the actual

dependent values? One rule to emphasize is that the regression model should

not be used to predict for independent variable values that are outside the range

of values used to construct the model. MINITAB issues a warning for such

activity when attempted. There are many instances where the relationship

between x and y are linear over a given interval but outside the interval the

relationship becomes curvilinear or unpredictable. Of course, with this caution

having been given, many forecasters use such regression models to extrapolate

to values of x outside the domain of those used to construct the model. Such

forecasts are introduced in section 14.8, “Using Regression to Develop a

Forecasting Trend Line”. Whether the forecasts obtained under such conditions
 P a g e | 404

are any better than "seat of the pants" or "crystal ball" estimates remains to be

seen.

The concept of residual analysis is a good one to show graphically and

numerically how the model relates to the data and the fact that it more
closely fits some points than others, etc. A graphical or numerical analysis of
residuals demonstrates that the regression line fits the data in a manner
analogous to the way a mean fits a set of numbers. The regression model
passes through the points such that the vertical distances from the actual y
values to the predicted values will sum to zero. The fact that the residuals
sum to zero points out the need to square the errors (residuals) in order to
get a handle on total error. This leads to the sum of squares error and then
on to the standard error of the estimate. In addition, students can learn why
the process is called least squares analysis (the slope and intercept formulas
are derived by calculus such that the sum of squares of error is minimized -
hence "least squares"). Students can learn that by examining the values of
se, the residuals, r2, and the t ratio to test the slope they can begin to make a
judgment about the fit of the model to the data. Many of the chapter
problems ask the student to comment on these items (se, r2, etc.).

It is my view that for many of these students, the most important facet
of this chapter lies in understanding the "buzz" words of regression such as
standard error of the estimate, coefficient of determination, etc. because they
may only interface regression again as some type of computer printout to be
deciphered. The concepts then may be more important than the calculations.
 P a g e | 405

CHAPTER OUTLINE

14.1 Introduction to Simple Regression Analysis

14.2 Determining the Equation of the Regression Line

14.3 Residual Analysis

Using Residuals to Test the Assumptions of the Regression Model

Using the Computer for Residual Analysis

14.4 Standard Error of the Estimate

14.5 Coefficient of Determination

Relationship Between r and r2

14.6 Hypothesis Tests for the Slope of the Regression Model and Testing the
Overall

Model

Testing the Slope

Testing the Overall Model

14.7 Estimation

Confidence Intervals to Estimate the Conditional Mean of y: µy/x

Prediction Intervals to Estimate a Single Value of y

 P a g e | 406

14.8 Using Regression to Develop a Forecasting Trend Line

Determining the Equation of the Trend Line

Forecasting Using the Equation of the Trend Line

Alternate Coding for Time Periods

14.8 Interpreting Computer Output

KEY TERMS

Coefficient of Determination (r2) Prediction Interval

Confidence Interval Probabilistic Model

Dependent Variable Regression Analysis

Deterministic Model Residual

Heteroscedasticity Residual Plot

Homoscedasticity Scatter Plot

Independent Variable Simple Regression

Least Squares Analysis Standard Error of the Estimate

(se)

Outliers Sum of Squares of Error (SSE)

 P a g e | 407

SOLUTIONS TO CHAPTER 14

14.1 x x

12 17

21 15

28 22

8 19

20 24

30

25

20

15
y

10

5 10 15 20 25 30

x = 89 y = 97 xy = 1,767

x2= 1,833 y2 = 1,935 n=5

 P a g e | 408

 x y (89)(97)
SS xy  xy  n
1,767 
 5
SS x ( x) 2 (89) 2
x 2

n
1,833 
5
b1 = = = 0.162

 y  b  x  97  0.162 89
1
n n 5 5
b0 = = 16.51

ŷ
= 16.51 + 0.162 x
 P a g e | 409

14.2 x_ _y_

140 25

119 29

103 46

91 70

65 88

29 112

24 128

x = 571 y = 498 xy = 30,099

x2 = 58,293 y2 = 45,154 n=7

 P a g e | 410

 x y (571)( 498)
SS xy  xy  n
30,099 
 7
SS x ( x) 2 (571) 2
x 2

n
58,293 
7
b1 = = = -0.898

 y  b  x  498  (0.898) 571

1
n n 7 7
b0 = = 144.414

ŷ
= 144.414 – 0.898 x
 P a g e | 411

14.3 (Advertising) x (Sales) y

12.5 148

3.7 55

21.6 338

60.0 994

37.6 541

6.1 89

16.8 126

41.2 379

x = 199.5 y = 2,670 xy = 107,610.4

x2 = 7,667.15 y2 = 1,587,328 n=8

 x y (199.5)( 2,670)
SS xy  xy  n
107,610.4 
 8
SS x ( x) 2 (199.5) 2
x 2

n
7,667.15 
8
b1 = = =
15.240

 y  b  x  2,670  15.24 199.5

1
n n 8 8
b0 = = -46.292

ŷ
= -46.292 + 15.240 x
 P a g e | 412

14.4 (Prime) x (Bond) y

16 5

6 12

8 9

4 15

7 7

x = 41 y = 48 xy = 333

x2 = 421 y2 = 524 n=5

 x y ( 41)( 48)
SS xy  xy  n
333 
 5
SS x ( x) 2 (41) 2
x 2

n
421 
5
b1 = = = -0.715

 y  b  x  48  (0.715) 41
1
n n 5 5
b0 = = 15.460

ŷ
= 15.460 – 0.715 x

14.5 Bankruptcies(y) Firm Births(x)

34.3 58.1
 P a g e | 413

35.0 55.4

38.5 57.0

40.1 58.5

35.5 57.4

37.9 58.0

x = 344.4 y = 221.3 x2 = 19,774.78

y2 = 8188.41 xy = 12,708.08 n=6

 x y (344.4)( 221.3)
SS xy  xy  n
12,708.08 
6

SS x ( x) 2 (344.4) 2
x 2

n
19,774.78 
6
b1 = = =

b1 = 0.878

 y  b  x  221.3  (0.878) 344.4

1
n n 6 6
b0 = = -13.503

ŷ
= -13.503 + 0.878 x
 P a g e | 414

14.6 No. of Farms (x) Avg. Size (y)

5.65 213

4.65 258

3.96 297

3.36 340

2.95 374

2.52 420

2.44 426

2.29 441

2.15 460

2.07 469

2.17 434

2.10 444

x = 36.31 y = 4,576 x2 = 124.7931

y2 = 1,825,028 xy = 12,766.71 n = 12

 x y (36.31)( 4,576)
SS xy  xy  n
12,766.71 
12

SS x ( x) 2 (36.31) 2
x 2

n
124.7931 
12
b1 = = = -72.328
 P a g e | 415

 y  b  x  4,576  (72.328) 36.31

1
n n 12 12
b0 = = 600.186

ŷ
= 600.186 – 72.3281 x
 P a g e | 416

14.7 Steel New Orders

99.9 2.74

97.9 2.87

98.9 2.93

87.9 2.87

92.9 2.98

97.9 3.09

100.6 3.36

104.9 3.61

105.3 3.75

108.6 3.95

x = 994.8 y = 32.15 x2 = 99,293.28

y2 = 104.9815 xy = 3,216.652 n = 10

 x y (994.8)(32.15)
SS xy  xy  n
3,216.652 
 10
SS x ( x) 2 (994.8) 2
x 2

n
99,293.28 
10
b1 = = =
0.05557
 P a g e | 417

 y  b  x  32.15  (0.05557) 994.8

1
n n 10 10
b0 = = -2.31307

ŷ
= -2.31307 + 0.05557 x
 P a g e | 418

14.8 x y

15 47

8 36

19 56

12 44

5 21

ŷ
= 13.625 + 2.303 x

Residuals:

ŷ ŷ
x y Residuals (y- )

15 47 48.1694 -1.1694

8 36 32.0489 3.9511

19 56 57.3811 -1.3811
12 44 41.2606 2.7394

5 21 25.1401 -4.1401

ŷ ŷ
14.9 x y Predicted ( ) Residuals (y- )

12 17 18.4582 -1.4582
 P a g e | 419

21 15 19.9196 -4.9196

28 22 21.0563 0.9437
8 19 17.8087 1.1913

20 24 19.7572 4.2428

ŷ
= 16.51 + 0.162 x

ŷ ŷ
14.10 x y Predicted ( ) Residuals (y- )

140 25 18.6597 6.3403

119 29 37.5229 -8.5229

103 46 51.8948 -5.8948

91 70 62.6737 7.3263

65 88 86.0281 1.9720

29 112 118.3648 -6.3648

24 128 122.8561 5.1439

ŷ
= 144.414 - 0.898 x
 P a g e | 420

ŷ ŷ
14.11 x y Predicted ( ) Residuals (y- )

12.5 148 144.2053 3.7947

3.7 55 10.0954 44.9047

21.6 338 282.8873 55.1127
60.0 994 868.0945 125.9055

37.6 541 526.7236 14.2764

6.1 89 46.6708 42.3292

16.8 126 209.7364 -83.7364

41.2 379 581.5868 -202.5868

ŷ
= -46.292 + 15.240x
 P a g e | 421

ŷ ŷ
14.12 x y Predicted ( ) Residuals (y- )

16 5 4.0259 0.9741

6 12 11.1722 0.8278
8 9 9.7429 -0.7429

4 15 12.6014 2.3986

7 7 10.4576 -3.4575

ŷ
= 15.460 - 0.715 x

ŷ ŷ
14.13 _ x_ _ y_ Predicted ( ) Residuals (y- )

58.1 34.3 37.4978 -3.1978

55.4 35.0 35.1277 -0.1277

57.0 38.5 36.5322 1.9678

58.5 40.1 37.8489 2.2511

57.4 35.5 36.8833 -1.3833

58.0 37.9 37.4100 0.4900

The residual for x = 58.1 is relatively large, but the residual for x =
55.4 is quite
 P a g e | 422

ŷ ŷ
small. 14.14 _x_ _ y_ Predicted ( ) Residuals (y- )

5 47 42.2756 4.7244

7 38 38.9836 -0.9836
11 32 32.3997 -0.3996

12 24 30.7537 -6.7537

19 22 19.2317 2.7683

25 10 9.3558 0.6442

ŷ
= 50.506 - 1.646 x

No apparent violation of assumptions

ŷ ŷ
14.15 Miles (x) Cost y ( ) (y- )

1,245 2.64 2.5376 .1024

425 2.31 2.3322 -.0222

1,346 2.45 2.5629 -.1128

973 2.52 2.4694 .0506

 P a g e | 423

255 2.19 2.2896 -.0996

865 2.55 2.4424 .1076

1,080 2.40 2.4962 -.0962

296 2.37 2.2998 .0702

ŷ
= 2.2257 – 0.00025 x

No apparent violation of assumptions

14.16

Error terms appear to be non independent

 P a g e | 424

14.17

There appears to be nonlinear regression

14.18 The MINITAB Residuals vs. Fits graphic is strongly indicative of a

violation of the homoscedasticity assumption of regression. Because
the residuals are very close together for small values of x, there is little
variability in the residuals at the left end of the graph. On the other
hand, for larger values of x, the graph flares out indicating a much
greater variability at the upper end. Thus, there is a lack of
homogeneity of error across the values of the independent variable.
 P a g e | 425

14.19 SSE = y2 – b0y - b1XY = 1,935 - (16.51)(97) - 0.1624(1767) =

46.5692

SSE 46.5692
se  
n2 3
= 3.94

Approximately 68% of the residuals should fall within ±1se.

3 out of 5 or 60% of the actual residuals fell within ± 1se.

14.20 SSE = y2 – b0y - b1XY = 45,154 - 144.414(498) - (-.89824)(30,099)

=

SSE = 272.0

SSE 272.0
se  
n2 5
= 7.376

6 out of 7 = 85.7% fall within + 1se

7 out of 7 = 100% fall within + 2se

 P a g e | 426

14.21 SSE = y2 – b0y - b1XY = 1,587,328 - (-46.29)(2,670) -

15.24(107,610.4) =

SSE = 70,940

SSE 70,940
se  
n2 6
= 108.7

Six out of eight (75%) of the sales estimates are within $108.7 million.

14.22 SSE = y2 – b0y - b1XY = 524 - 15.46(48) - (-0.71462)(333) =

19.8885

SSE 19.8885
se  
n2 3
= 2.575

Four out of five (80%) of the estimates are within 2.575 of the actual
rate for

bonds. This amount of error is probably not acceptable to financial

analysts.
 P a g e | 427

ŷ ŷ ( y  yˆ ) 2
14.23 _ x_ _ y_ Predicted ( ) Residuals (y- )

58.1 34.3 37.4978 -3.1978 10.2259

55.4 35.0 35.1277 -0.1277 0.0163
57.0 38.5 36.5322 1.9678 3.8722

58.5 40.1 37.8489 2.2511 5.0675

57.4 35.5 36.8833 -1.3833 1.9135

58.0 37.9 37.4100 0.4900 0.2401

 ( y  yˆ ) 2

= 21.3355

 ( y  yˆ ) 2

SSE = = 21.3355

SSE 21.3355
se  
n2 4
= 2.3095

This standard error of the estimate indicates that the regression

model is
with + 2.3095(1,000) bankruptcies about 68% of the time. In
this

particular problem, 5/6 or 83.3% of the residuals are within this

standard

error of the estimate.

 P a g e | 428

ŷ ŷ
14.24 (y- ) (y- )2

4.7244 22.3200

-0.9836 .9675

-0.3996 .1597

-6.7537 45.6125

2.7683 7.6635
0.6442 .4150

ŷ
(y- )2 = 77.1382

 ( y  yˆ ) 2

SSE = = 77.1382

SSE 77.1382
se  
n2 4
= 4.391
 P a g e | 429

ŷ ŷ
14.25 (y- ) (y- )2

.1024 .0105

-.0222 .0005

-.1129 .0127

.0506 .0026
-.0996 .0099

.1076 .0116

-.0962 .0093

.0702 .0049

ŷ  ( y  yˆ ) 2

(y- )2 = .0620 SSE = =

.0620

SSE .0620
se  
n2 6
= .1017

The model produces estimates that are ±.1017 or within about 10

cents 68% of the

time. However, the range of milk costs is only 45 cents for this data.

14.26 Volume (x) Sales (y)

728.6 10.5

497.9 48.1

439.1 64.8
 P a g e | 430

377.9 20.1

375.5 11.4

363.8 123.8

276.3 89.0

n=7 x = 3059.1 y = 367.7

x2 = 1,464,071.97 y2 = 30,404.31 xy = 141,558.6

b1 = -.1504 b0 = 118.257

ŷ
= 118.257 - .1504x

SSE = y2 – b0y - b1XY

= 30,404.31 - (118.257)(367.7) - (-0.1504)(141,558.6) =

8211.6245

SSE 8211.6245
se  
n2 5
= 40.526

This is a relatively large standard error of the estimate given the sales
values

(ranging from 10.5 to 123.8).

 P a g e | 431

SSE 46.6399
1  1
( y ) 2
(97) 2
y  n
2
1,935 
5
14.27 r2 = = .123

This is a low value of r2

SSE 272.121
1 1
( y ) 2
(498) 2
y  n
2
45,154 
7
14.28 r2 = = .972

This is a high value of r2

SSE 70,940
1 1
( y ) 2
(2,670) 2
 y 2

n
1,587,328 
8
14.29 r2 = = .898

This value of r2 is relatively high

SSE 19.8885
1  1
( y ) 2
(48) 2
y  n
2
524 
5
14.30 r2 = = .685
 P a g e | 432

This value of r2 is a modest value.

68.5% of the variation of y is accounted for by x but 31.5% is

unaccounted for.

SSE 21.33547
1 1
( y ) 2
(221.3) 2
y  n
2
8,188.41 
6
14.31 r2 = = .183

This value is a low value of r2.

Only 18.3% of the variability of y is accounted for by the x values and

81.7% are

unaccounted for.

14.32 CCI Median Income

116.8 37.415

91.5 36.770

68.5 35.501

61.6 35.047

65.9 34.700

90.6 34.942

100.0 35.887
 P a g e | 433

104.6 36.306

125.4 37.005

x = 323.573 y = 824.9 x2 = 11,640.93413

y2 = 79,718.79 xy = 29,804.4505 n=9

 x y (323.573)(824.9)
SS xy  xy  n
29,804.4505 
 9
SS x ( x) 2 (323.573) 2
x 2

n
11,640.93413 
9
b1 = = =

b1 = 19.2204

 y  b  x  824.9  (19.2204) 323.573

1
n n 9 9
b0 = = -599.3674

ŷ
= -599.3674 + 19.2204 x

SSE = y2 – b0y - b1XY =

79,718.79 – (-599.3674)(824.9) – 19.2204(29,804.4505) =

1283.13435

SSE 1283.13435
se  
n2 7
= 13.539
 P a g e | 434

SSE 1283.13435
1  1
( y ) 2
(824.9) 2
y  n
2
79,718.79 
9
r2 = = .688

se 3.94

( x ) 2 (89) 2
 x2  n
1.833 
5
14.33 sb = = .2498

b1 = 0.162

Ho:  = 0  = .05

Ha:   0

This is a two-tail test, /2 = .025 df = n - 2 = 5 - 2 = 3

t.025,3 = ±3.182

b1   1 0.162  0

sb .2498
t = = 0.65
 P a g e | 435

Since the observed t = 0.65 < t.025,3 = 3.182, the decision is to fail to
reject the

null hypothesis.

se 7.376

( x ) 2 (571) 2
 x2  n
58,293 
7
14.34 sb = = .068145

b1 = -0.898

Ho:  = 0  = .01

Ha:   0

Two-tail test, /2 = .005 df = n - 2 = 7 - 2 = 5

t.005,5 = ±4.032

b1   1  0.898  0

sb .068145
t= = -13.18

Since the observed t = -13.18 < t.005,5 = -4.032, the decision is to

reject the null

hypothesis.
 P a g e | 436

se 108.7

( x ) 2 (199.5) 2
 x2  n
7,667.15 
8
14.35 sb = = 2.095

b1 = 15.240

Ho:  = 0  = .10

Ha:   0

For a two-tail test, /2 = .05 df = n - 2 = 8 - 2 = 6

t.05,6 = 1.943

b1   1 15,240  0

sb 2.095
t = = 7.27

Since the observed t = 7.27 > t.05,6 = 1.943, the decision is to reject
the null

hypothesis.
 P a g e | 437

se 2.575

( x ) 2
(41) 2
x 2

n
421 
5
14.36 sb = = .27963

b1 = -0.715

Ho:  = 0  = .05

Ha:   0

For a two-tail test, /2 = .025 df = n - 2 = 5 - 2 = 3

t.025,3 = ±3.182

b1   1  0.715  0

sb .27963
t = = -2.56

Since the observed t = -2.56 > t.025,3 = -3.182, the decision is to fail to
reject the

null hypothesis.
 P a g e | 438

se 2.3095

( x ) 2
(344.4) 2
x 2

n
19,774.78 
6
14.37 sb = = 0.926025

b1 = 0.878

Ho:  = 0  = .05

Ha:   0

For a two-tail test, /2 = .025 df = n - 2 = 6 - 2 = 4

t.025,4 = ±2.776

b1  1 0.878  0

sb .926025
t = = 0.948

Since the observed t = 0.948 < t.025,4 = 2.776, the decision is to fail to
reject the

null hypothesis.

14.38 F = 8.26 with a p-value of .021. The overall model is significant at 

= .05 but
 P a g e | 439

not at  = .01. For simple regression,

F
t = = 2.874

t.05,8 = 1.86 but t.01,8 = 2.896. The slope is significant at  = .05 but not
at

 = .01.
 P a g e | 440

14.39 x0 = 25

95% confidence /2 = .025

df = n - 2 = 5 - 2 = 3 t.025,3 = ±3.182

x
 x  89
n 5
= 17.8

x = 89 x2 = 1,833

se = 3.94

ŷ
= 16.5 + 0.162(25) = 20.55

1 ( x 0  x) 2

n ( x ) 2
ŷ  x 2

n
± t /2,n-2 se

1 (25  17.8) 2

5 (89) 2
1,833 
5
20.55 ± 3.182(3.94) = 20.55 ± 3.182(3.94)
(.63903) =

20.55 ± 8.01
 P a g e | 441

12.54 < E(y25) < 28.56

 P a g e | 442

14.40 x0 = 100 For 90% confidence, /2 = .05

df = n - 2 = 7 - 2 = 5 t.05,5 = ±2.015

x
 x  571
n 7
= 81.57143

x= 571 x2 = 58,293 se = 7.377

ŷ
= 144.414 - .0898(100) = 54.614

1 ( x0  x ) 2
1 
n ( x ) 2
ŷ x  n
2

± t /2,n-2 se =

1 (100  81.57143) 2
1 
7 (571) 2
58,293 
7
54.614 ± 2.015(7.377) =

54.614 ± 2.015(7.377)(1.08252) = 54.614 ± 16.091

38.523 < y < 70.705

 P a g e | 443

ŷ
For x0 = 130, = 144.414 - 0.898(130) = 27.674

1 ( x0  x) 2
1 
n ( x ) 2
x  n
2

y ± t /2,n-2 se =

1 (130  81.57143) 2
1 
7 (571) 2
58,293 
7
27.674 ± 2.015(7.377) =

27.674 ± 2.015(7.377)(1.1589) = 27.674 ± 17.227

10.447 < y < 44.901

The width of this confidence interval of y for x0 = 130 is wider that the

confidence interval of y for x0 = 100 because x0 = 100 is nearer to the

value of

x = 81.57 than is x0 = 130.

14.41 x0 = 20 For 98% confidence, /2 = .01

df = n - 2 = 8 - 2 = 6 t.01,6 = 3.143

x
 x  199.5
n 8
= 24.9375
 P a g e | 444

x = 199.5 x2 = 7,667.15 se = 108.8

ŷ
= -46.29 + 15.24(20) = 258.51

1 ( x 0  x) 2

n ( x ) 2
ŷ  x 2

n
± t /2,n-2 se

1 (20  24.9375) 2

8 (199.5) 2
7,667.15 
8
258.51 ± (3.143)(108.8)

258.51 ± (3.143)(108.8)(0.36614) = 258.51 ± 125.20

133.31 < E(y20) < 383.71

For single y value:

1 ( x0  x) 2
1 
n ( x ) 2
ŷ x  n
2

± t /2,n-2 se
 P a g e | 445

1 (20  24.9375) 2
1 
8 (199.5) 2
7,667.15 
8
258.51 ± (3.143)(108.8)

258.51 ± (3.143)(108.8)(1.06492) = 258.51 ± 364.16

-105.65 < y < 622.67

The confidence interval for the single value of y is wider than the
confidence

interval for the average value of y because the average is more

towards the

middle and individual values of y can vary more than values of the
average.
 P a g e | 446

14.42 x0 = 10 For 99% confidence /2 = .005

df = n - 2 = 5 - 2 = 3 t.005,3 = 5.841

x
 x  41
n 5
= 8.20

x = 41 x2 = 421 se = 2.575

ŷ
= 15.46 - 0.715(10) = 8.31

1 ( x 0  x) 2

n ( x ) 2
ŷ x  n
2

± t /2,n-2 se

1 (10  8.2) 2

5 (41) 2
421 
5
8.31 ± 5.841(2.575) =

8.31 ± 5.841(2.575)(.488065) = 8.31 ± 7.34

0.97 < E(y10) < 15.65

 P a g e | 447

If the prime interest rate is 10%, we are 99% confident that the
average bond rate

is between 0.97% and 15.65%.

 P a g e | 448

14.43 Year Fertilizer

2001 11.9
2002 17.9

2003 22.0

2004 21.8

2005 26.0

x = 10,015 y = 99.6 xy = 199,530.9

x2= 20,060,055 y2 = 2097.26 n=5

 x y (10,015)(99.6)
SS xy  xy  n
199,530.9 
5

SS x ( x) 2 (10,015) 2
x 2

n
20,060,055 
5
b1 = = = 3.21

 y  b  x  99.6  3.2110,015
1
n n 5 5
b0 = = -6,409.71

ŷ
= -6,409.71 + 3.21 x

ŷ
(2008) = -6,409.71 + 3.21(2008) = 35.97
 P a g e | 449

14.44 Year Fertilizer

1998 5860
1999 6632
2000 7125
2001 6000
2002 4380

2003 3326

2004 2642

x = 14,007 y = 35,965 xy = 71,946,954

x2= 28,028,035 y2 = 202,315,489 n=7

 x y (14,007)(35,965)
SS xy  xy  n
71,946,954 
7

SS x ( x) 2 (14,007) 2
x 2

n
28,028,035 
7
b1 = = =
-678.9643

 y  b  x  35,965  678.9643 14,007

1
n n 7 7
b0 = = 1,363,745.39

ŷ
= 1,363,745.39 + -678.9643 x

ŷ
(2007) = 1,363,745.39 + -678.9643(2007) = 1,064.04
 P a g e | 450

14.45 Year Quarter Cum. Quarter(x) Sales(y)

2003 1 1 11.93

2 2 12.46

3 3 13.28

4 4 15.08

2004 1 5 16.08

2 6 16.82

3 7 17.60

4 8 18.66

2005 1 9 19.73

2 10 21.11

3 11 22.21

4 12 22.94

Use the cumulative quarters as the predictor variable, x, to predict

sales, y.

x = 78 y = 207.9 xy = 1,499.07

x2= 650 y2 = 3,755.2084 n = 12

 x y (78)( 207.9)
SS xy  xy  n
1,499.07 
12

SS x ( x) 2 (78) 2
x 2

n
650 
12
b1 = = = 1.033
 P a g e | 451

 y  b  x  207.9  1.033 78
1
n n 12 12
b0 = = 10.6105

ŷ
= 10.6105 + 1.033 x

Remember, this trend line was constructed using cumulative quarters.

To forecast

sales for the third quarter of year 2007, we must convert this time
frame to

cumulative quarters. The third quarter of year 2007 is quarter number

19 in our

scheme.

ŷ
(19) = 10.6105 + 1.033(19) = 30.2375
 P a g e | 452

14.46 x y

5 8

7 9

3 11

16 27

12 15

9 13

x = 52 x2 = 564

y = 83 y2 = 1,389 b1 = 1.2853

xy = 865 n=6 b0 = 2.6941

ŷ
a) = 2.6941 + 1.2853 x

ŷ ŷ
b) (Predicted Values) (y- ) residuals

9.1206 -1.1206

11.6912 -2.6912

6.5500 4.4500

23.2588 3.7412

18.1177 -3.1176

14.2618 -1.2618
 P a g e | 453

ŷ
c) (y- )2

1.2557

7.2426

19.8025

13.9966

9.7194

1.5921

SSE = 53.6089

SSE 53.6089
se  
n2 4
= 3.661

SSE 53.6089
1  1
( y ) 2
(83) 2
y  n
2
1,389 
6
d) r2 = = .777
 P a g e | 454

e) Ho:  = 0  = .01

Ha:   0

Two-tailed test, /2 = .005 df = n - 2 = 6 - 2 = 4

t.005,4 = ±4.604

se 3.661

( x ) 2 (52) 2
x 2

n
564 
6
sb = = .34389

b1   1 1.2853  0

sb .34389
t = = 3.74

Since the observed t = 3.74 < t.005,4 = 4.604, the decision is to

fail to reject

the null hypothesis.

f) The r2 = 77.74% is modest. There appears to be some

prediction with this

model. The slope of the regression line is not significantly

different from
 P a g e | 455

zero using  = .01. However, for  = .05, the null hypothesis of

a zero

slope is rejected. The standard error of the estimate, se = 3.661

is not

particularly small given the range of values for y (11 - 3 = 8).

14.47 x y

53 5

47 5

41 7

50 4

58 10

62 12

45 3

60 11

x = 416 x2 = 22,032

y = 57 y2 = 489 b1 = 0.355

xy = 3,106 n=8 b0 = -11.335

ŷ
a) = -11.335 + 0.355 x
 P a g e | 456

ŷ ŷ
b) (Predicted Values) (y- ) residuals

7.48 -2.48

5.35 -0.35

3.22 3.78

6.415 -2.415

9.255 0.745

10.675 1.325

4.64 -1.64

9.965 1.035

ŷ
c) (y- )2

6.1504

0.1225

14.2884

5.8322

0.5550

1.7556

2.6896

1.0712

SSE = 32.4649

SSE 32.4649

n2 6
d) se = = 2.3261
 P a g e | 457

SSE 32.4649
1  1
( y ) 2
(57) 2
y  n
2
489 
8
e) r2 = = .608

f) Ho:  = 0  = .05

Ha:   0

Two-tailed test, /2 = .025 df = n - 2 = 8 - 2 = 6

t.025,6 = ±2.447

se 2.3261

( x ) 2 (416) 2
 x2  n
22,032 
8
sb = = 0.116305

b1   1 0.3555  0

sb .116305
t = = 3.05

Since the observed t = 3.05 > t.025,6 = 2.447, the decision is to

reject the

null hypothesis.

The population slope is different from zero.

 P a g e | 458

g) This model produces only a modest r2 = .608. Almost 40% of

the

variance of y is unaccounted for by x. The range of y values is

12 - 3 = 9

and the standard error of the estimate is 2.33. Given this

small range, the

se is not small.

14.48 x = 1,263 x2 = 268,295

y = 417 y2 = 29,135

xy = 88,288 n=6

b0 = 25.42778 b1 = 0.209369

SSE = y2 - b0y - b1xy =

29,135 - (25.42778)(417) - (0.209369)(88,288) = 46.845468

SSE 46.845468
1  1
( y ) 2
153.5
 y2  n
r2 = = .695
 P a g e | 459

Coefficient of determination = r2 = .695

14.49a) x0 = 60

x = 524 x2 = 36,224

y = 215 y2 = 6,411 b1 = .5481

xy = 15,125 n=8 b0 =

-9.026

se = 3.201 95% Confidence Interval /2 = .025

df = n - 2 = 8 - 2 = 6

t.025,6 = ±2.447

ŷ
= -9.026 + 0.5481(60) = 23.86

x
 x  524
n 8
= 65.5
 P a g e | 460

1 ( x0  x) 2

n ( x ) 2
ŷ x  n
2

± t /2,n-2 se

1 (60  65.5) 2

8 (524) 2
36,224 
8
23.86 + 2.447(3.201)

23.86 + 2.447(3.201)(.375372) = 23.86 + 2.94

20.92 < E(y60) < 26.8

b) x0 = 70

ŷ
70 = -9.026 + 0.5481(70) = 29.341

1 ( x 0  x) 2
1 
n ( x ) 2
ŷ x  n
2

+ t/2,n-2 se

1 (70  65.5) 2
1 
8 (524) 2
36,224 
8
29.341 + 2.447(3.201)
 P a g e | 461

29.341 + 2.447(3.201)(1.06567) = 29.341 + 8.347

20.994 < y < 37.688

c) The confidence interval for (b) is much wider because part (b) is for a
single value

of y which produces a much greater possible variation. In actuality, x0

= 70 in

part (b) is slightly closer to the mean (x) than x0 = 60. However, the
width of the

single interval is much greater than that of the average or expected y

value in

part (a).
 P a g e | 462

14.50 Year Cost

1 56
2 54

3 49

4 46

5 45

x = 15 y = 250 xy = 720

x2= 55 y2 = 12,594 n=5

 x y (15)( 250)
SS xy  xy  n
720 
5

SS x ( x) 2 (15) 2
x 2

n
55 
5
b1 = = = -3

 y  b  x  250  (3) 15
1
n n 5 5
b0 = = 59

ŷ
= 59 - 3 x

ŷ
(7) = 59 - 3(7) = 38
 P a g e | 463

14.51 y = 267 y2 = 15,971

x = 21 x2 = 101

xy = 1,256 n =5

b0 = 9.234375 b1 = 10.515625

SSE = y2 - b0y - b1xy =

15,971 - (9.234375)(267) - (10.515625)(1,256) = 297.7969

SSE 297.7969
1  1
( y ) 2
1,713.2
 y2  n
r2 = = .826

If a regression model would have been developed to predict number of

cars sold

by the number of sales people, the model would have had an r2 of

82.6%. The

same would hold true for a model to predict number of sales people by
the

number of cars sold.

14.52 n = 12 x = 548 x2 = 26,592

y = 5940 y2 = 3,211,546 xy = 287,908

 P a g e | 464

b1 = 10.626383 b0 = 9.728511

ŷ
= 9.728511 + 10.626383 x

SSE = y2 - b0y - b1xy =

3,211,546 - (9.728511)(5940) - (10.626383)(287,908) = 94337.9762

SSE 94,337.9762
se  
n2 10
= 97.1277

SSE 94,337.9762
1  1
( y ) 2
271,246
 y2  n
r2 = = .652

10.626383  0
97.1277
(548) 2
26,592 
12
t = = 4.33

If  = .01, then t.005,10 = 3.169. Since the observed t = 4.33 > t.005,10 =
3.169, the

decision is to reject the null hypothesis.

P a g e | 465
 P a g e | 466

14.53 Sales(y) Number of Units(x)

17.1 12.4

7.9 7.5

4.8 6.8

4.7 8.7

4.6 4.6

4.0 5.1

2.9 11.2

2.7 5.1

2.7 2.9

y = 51.4 y2 = 460.1 x = 64.3

x2 = 538.97 xy = 440.46 n=9

b1 = 0.92025 b0 = -0.863565

ŷ
= -0.863565 + 0.92025 x

SSE = y2 - b0y - b1xy =

460.1 - (-0.863565)(51.4) - (0.92025)(440.46) = 99.153926

 P a g e | 467

SSE 99.153926
1  1
( y ) 2
166.55
 y 2

n
r2 = = .405
 P a g e | 468

14.54 Year Total Employment

1995 11,152

1996 10,935

1997 11,050

1998 10,845

1999 10,776

2000 10,764

2001 10,697

2002 9,234

2003 9,223

2004 9,158

x = 19,995 y = 103,834 xy =

207,596,350

x2= 39,980,085 y2 = 1,084,268,984 n=7

 x y (19,995)(103,834)
SS xy  xy  n
207,596,350 
10

SS x ( x) 2 (19,995) 2
x 2

n
39,980,085 
10
b1 = = = -239.188

 y  b  x  103,834  (239.188) 19,995

1
n n 10 10
b0 = = 488,639.564
 P a g e | 469

ŷ
= 488,639.564 + -239.188 x

ŷ
(2008) = 488,639.564 + -239.188(2008) = 8,350.30
 P a g e | 470

14.55 1977 2003

581 666

213 214

668 496

345 204

1476 1600

1776 6278

x= 5059 y = 9458 x2 = 6,280,931

y2 = 42,750,268 xy = 14,345,564 n = 6

 x y (5059)(9358)
SS xy  xy  n
13,593,272 
6

SS x ( x) 2 (5059) 2
x 2

n
6,280,931 
6
b1 = = = 3.1612

 y  b  x  9458  (3.1612) 5059

1
n n 6 6
b0 = = -1089.0712

ŷ
= -1089.0712 + 3.1612 x
 P a g e | 471

for x = 700:

ŷ
= 1076.6044

1 ( x 0  x) 2

n ( x ) 2
ŷ x  n
2

+ t/2,n-2se

 = .05, t.025,4 = 2.776

x0 = 700, n = 6

x
= 843.167

SSE = y2 – b0y –b1xy =

42,750,268 – (-1089.0712)(9458) – (3.1612)(14,345,564) =

7,701,506.49

SSE 7,701,506.49
se  
n2 4
= 1387.58

Confidence Interval =
 P a g e | 472

1 (700  843.167) 2

6 (5059) 2
6,280,931 
6
1123.757 + (2.776)(1387.58) =

1123.757 + 1619.81

-496.05 to 2743.57

H0: 1 = 0

Ha: 1  0

 = .05 df = 4

Table t.025,4 = 2.776

b1  0 3.1612  0 2.9736
 
sb 1387.58 .8231614
2,015,350.833
t = = 3.234

Since the observed t = 3.234 > t.025,4 = 2.776, the decision is to reject
the null

hypothesis.

14.56 x = 11.902 x2 = 25.1215

y = 516.8 y2 = 61,899.06 b1 = 66.36277

 P a g e | 473

xy = 1,202.867 n=7 b0 = -39.0071

ŷ
= -39.0071 + 66.36277 x

SSE = y2 - b0 y - b1 xy

SSE = 61,899.06 - (-39.0071)(516.8) - (66.36277)(1,202.867) =

2,232.343

SSE 2,232.343
se  
n2 5
= 21.13

SSE 2,232.343
1 1
( y ) 2
(516.8) 2
 y 2

n
61,899.06 
7
r2 = = 1 - .094 = .906

14.57 x = 44,754 y = 17,314 x2 = 167,540,610

y2 = 24,646,062 n = 13 xy = 59,852,571

 x y (44,754)(17,314)
SS xy  xy  n
59,852,571 
 13
SS x ( x) 2 (44,754) 2
x 2

n
167,540,610 
13
b1 = = = .01835
 P a g e | 474

 y  b  x  17,314  (.01835) 44,754

1
n n 13 13
b0 = = 1268.685

ŷ
= 1268.685 + .01835 x

r2 for this model is .002858. There is no predictability in this model.

Test for slope: t = 0.18 with a p-value of 0.8623. Not significant

Time-Series Trend Line:

x = 91 y = 44,754 xy = 304,797

x2= 819 y2 = 167,540,610 n = 13

 x y (91)( 44,754)
SS xy  xy  n
304,797 
13

SS x ( x) 2 (91) 2
x 2

n
819 
13
b1 = = = -46.5989

 y  b  x  44,754  (46.5989) 91
1
n n 13 13
b0 = = 3,768.81

ŷ
= 3,768.81 – 46.5989 x
 P a g e | 475

ŷ
(2007) = 3,768.81 - 46.5989(15) = 3,069.83
 P a g e | 476

14.58 x = 323.3 y = 6765.8

x2 = 29,629.13 y2 = 7,583,144.64

xy = 339,342.76 n = 7

 x y (323.3)(6765.8)
SS xy  xy  n
339,342.76 
7

SS x ( x) 2 (323.3) 2
x 2

n
29,629.13 
7
b1 = = = 1.82751

 y  b  x  6765.8  (1.82751) 323.3

1
n n 7 7
b0 = = 882.138

ŷ
= 882.138 + 1.82751 x

SSE = y2 –b0y –b1xy

= 7,583,144.64 –(882.138)(6765.8) –(1.82751)(339,342.76) =

994,623.07

SSE 994,623.07
se  
n2 5
= 446.01
 P a g e | 477

SSE 994,623.07
1  1
( y ) 2
(6765.8) 2
y  n
2
7,583,144 .64 
7
r2 = = 1 - .953 = .
047

H 0:  = 0

Ha:   0  = .05 t.025,5 = 2.571

  x 2
(323.3) 2
x  2

n
 29,629.13 
7
SSxx = = 14,697.29

b1  0 1.82751  0

se 446.01
SS xx 14,697.29
t = = 0.50

Since the observed t = 0.50 < t.025,5 = 2.571, the decision is to fail to
reject the

null hypothesis.
 P a g e | 478

14.59 Let Water use = y and Temperature = x

x = 608 x2 = 49,584

y = 1,025 y2 = 152,711 b1 = 2.40107

xy = 86,006 n=8 b0 = -54.35604

ŷ
= -54.35604 + 2.40107 x

ŷ
100 = -54.35604 + 2.40107(100) = 185.751

SSE = y2 - b0 y - b1 xy

SSE = 152,711 - (-54.35604)(1,025) - (2.40107)(86,006) = 1919.5146

SSE 1,919.5146
se  
n2 6
= 17.886

SSE 1,919.5145
1  1
( y ) 2
(1025) 2
y  n
2
152,711 
8
r2 = = 1 - .09 = .91

Testing the slope:

 P a g e | 479

Ho:  = 0

Ha:   0  = .01

Since this is a two-tailed test, /2 = .005

df = n - 2 = 8 - 2 = 6

t.005,6 = ±3.707

se 17.886

( x) 2 (608) 2
x 2

n
49,584 
8
sb = = .30783

b1   1 2.40107  0

sb .30783
t = = 7.80

Since the observed t = 7.80 < t.005,6 = 3.707, the decision is to reject
the null

hypothesis.

ŷ
14.60 a) The regression equation is: = 67.2 – 0.0565 x

b) For every unit of increase in the value of x, the predicted value of y

will
 P a g e | 480

decrease by -.0565.

c) The t ratio for the slope is –5.50 with an associated p-value of .000.
This is

significant at  = .10. The t ratio negative because the slope is

negative and

the numerator of the t ratio formula equals the slope minus zero.

d) r2 is .627 or 62.7% of the variability of y is accounted for by x. This

is only

a modest proportion of predictability. The standard error of the

estimate is

10.32. This is best interpreted in light of the data and the

magnitude of the

data.

e) The F value which tests the overall predictability of the model is

30.25. For

simple regression analysis, this equals the value of t2 which is (-

5.50)2.

f) The negative is not a surprise because the slope of the regression

line is also

negative indicating an inverse relationship between x and y. In

addition,

taking the square root of r2 which is .627 yields .7906 which is the
magnitude

of the value of r considering rounding error.

 P a g e | 481

14.61 The F value for overall predictability is 7.12 with an associated p-value
of .0205 which is significant at  = .05. It is not significant at alpha of .
01. The coefficient of determination is .372 with an adjusted r2 of .
32. This represents very modest predictability. The standard error of
the estimate is 982.219, which in units of 1,000 laborers means that
about 68% of the predictions are within 982,219 of the actual figures.
The regression model is:

Number of Union Members = 22,348.97 - 0.0524 Labor Force. For a

labor force of 100,000 (thousand, actually 100 million), substitute x =
100,000 and get a predicted value of 17,108.97 (thousand) which is
actually 17,108,970 union members.

14.62 The Residual Model Diagnostics from MINITAB indicate a relatively healthy
set
of residuals. The Histogram indicates that the error terms are
generally normally

distributed. This is somewhat confirmed by the semi straight line

Normal Plot of Residuals. However, the Residuals vs. Fits graph
indicates that there may be some heteroscedasticity with greater error
variance for small x values.
 P a g e | 482

Chapter 15
Multiple Regression Analysis

LEARNING OBJECTIVES

This chapter presents the potential of multiple regression analysis as a tool in

business decision making and its applications, thereby enabling you to:

1. Develop a multiple regression model.

2. Understand and apply significance tests of the regression model and

its coefficients.
 P a g e | 483

3. Compute and interpret residuals, the standard error of the estimate,

and the coefficient of determination.

4. Interpret multiple regression computer output.

CHAPTER TEACHING STRATEGY

In chapter 14 using simple regression, the groundwork was prepared

for chapter 15 by presenting the regression model along with mechanisms for
testing the strength of the model such as se, r2, a t test of the slope, and the
residuals. In this chapter, multiple regression is presented as an extension of
the simple linear regression case. It is initially pointed out that any model
that has at least one interaction term or a variable that represents a power of
two or more is considered a multiple regression model. Multiple regression
opens up the possibilities of predicting by multiple independent variables and
nonlinear relationships. It is emphasized in the chapter that with both simple
and multiple regression models there is only one dependent variable. Where
simple regression utilizes only one independent variable, multiple regression
can utilize more than one independent variable.
 P a g e | 484

Presented early in chapter 15 are the simultaneous equations that

need to be solved to develop a first-order multiple regression model using
two predictors. This should help the student to see that there are three
equations with three unknowns to be solved. In addition, there are eight
values that need to be determined before solving the simultaneous equations
( x1, x2, y, x12, . . .) Suppose there are five predictors. Six simultaneous
equations must be solved and the number of sums needed as constants in
the equations become overwhelming. At this point, the student will begin to
realize that most researchers do not want to take the time nor the effort to
solve for multiple regression models by hand. For this reason, much of the
chapter is presented using computer printouts. The assumption is that the
use of multiple regression analysis is largely from computer analysis.

Topics included in this chapter are similar to the ones in chapter 14

including tests of the slope, R2, and se. In addition, an adjusted R2 is
introduced in chapter 15. The adjusted R2 takes into account the degrees of
freedom error and total degrees of freedom whereas R2 does not. If there is a
significant discrepancy between adjusted R2 and R2, then the regression
model may not be as strong as it appears to be with the R2. The gap between
R2 and adjusted R2 tends to increase as non significant independent variables
are added to the regression model and decreases with increased sample size.
 P a g e | 485

CHAPTER OUTLINE

15.1 The Multiple Regression Model

Multiple Regression Model with Two Independent Variables (First-Order)

Determining the Multiple Regression Equation

A Multiple Regression Model

15.2 Significant Tests of the Regression Model and its Coefficients

Testing the Overall Model

Significance Tests of the Regression Coefficients

15.3 Residuals, Standard Error of the Estimate, and R2

Residuals

SSE and Standard Error of the Estimate

Coefficient of Determination (R2)

Adjusted R2

15.4 Interpreting Multiple Regression Computer Output

A Reexamination of the Multiple Regression Output

 P a g e | 486

KEY TERMS

Adjusted R2 R2

Coefficient of Multiple Determination (R2) Residual

Dependent Variable

Response Plane
Independent Variable Response Surface

Least Squares Analysis Response Variable

Multiple Regression Standard Error of the Estimate

Outliers Sum of Squares of Error

Partial Regression Coefficient

SOLUTIONS TO PROBLEMS IN CHAPTER 15

15.1 The regression model is:

ŷ
= 25.03 - 0.0497 x1 + 1.928 x2
 P a g e | 487

Predicted value of y for x1 = 200 and x2 = 7 is:

ŷ
= 25.03 - 0.0497(200) + 1.928(7) = 28.586

15.2 The regression model is:

ŷ
= 118.56 - 0.0794 x1 - 0.88428 x2 + 0.3769 x3

Predicted value of y for x1 = 33, x2 = 29, and x3 = 13 is:

ŷ
= 118.56 - 0.0794(33) - 0.88428(29) + 0.3769(13) = 95.19538
 P a g e | 488

15.3 The regression model is:

ŷ
= 121.62 - 0.174 x1 + 6.02 x2 + 0.00026 x3 + 0.0041 x4

There are four independent variables. If x2, x3, and x4 are held
constant, the predicted y will decrease by - 0.174 for every unit
increase in x1. Predicted y will increase by 6.02 for every unit increase
in x2 as x1, x3, and x4 are held constant. Predicted y will increase by
0.00026 for every unit increase in x3 holding x1, x2, and x4 constant. If
x4 is increased by one unit, the predicted y will increase by 0.0041 if x1,
x2, and x3 are held constant.

15.4 The regression model is:

ŷ
= 31,409.5 + 0.08425 x1 + 289.62 x2 - 0.0947 x3

For every unit increase in x1, the predicted y increases by 0.08425 if x2

and x3 are held constant. The predicted y will increase by 289.62 for
every unit increase in x2 if x1 and x3 are held constant. The predicted y
will decrease by 0.0947 for every unit increase in x3 if x1 and x2 are
held constant.
 P a g e | 489

15.5 The regression model is:

Per Capita = -7,629.627 + 116.2549 Paper Consumption

- 120.0904 Fish Consumption + 45.73328 Gasoline

Consumption.

For every unit increase in paper consumption, the predicted per capita
consumption increases by 116.2549 if fish and gasoline consumption
are held constant. For every unit increase in fish consumption, the
predicted per capita consumption decreases by 120.0904 if paper and
gasoline consumption are held constant. For every unit increase in
gasoline consumption, the predicted per capita consumption increases
by 45.73328 if paper and fish consumption are held constant.
 P a g e | 490

15.6 The regression model is:

Insider Ownership =

17.68 - 0.0594 Debt Ratio - 0.118 Dividend Payout

The coefficients mean that for every unit of increase in debt ratio there
is a predicted decrease of - 0.0594 in insider ownership if dividend
payout is held constant. On the other hand, if dividend payout is
increased by one unit, then there is a predicted drop of insider
ownership by 0.118 with debt ratio is held constant.

15.7 There are 9 predictors in this model. The F test for overall significance
of the model is 1.99 with a probability of .0825. This model is not
significant at  = .05. Only one of the t values is statistically
significant. Predictor x1 has a t of 2.73 which has an associated
probability of .011 and this is significant at  = .05.

15.8 This model contains three predictors. The F test is significant at  = .

05 but not at  = .01. The t values indicate that only one of the three
predictors is significant. Predictor x1 yields a t value of 3.41 with an
associated probability of .005. The recommendation is to rerun the
model using only x1 and then search for other variables besides x2 and
x3 to include in future models.
 P a g e | 491

15.9 The regression model is:

Per Capita Consumption = -7,629.627 + 116.2549 Paper Consumption

- 120.0904 Fish Consumption + 45.73328 Gasoline Consumption

This model yields an F = 14.319 with p-value = .0023. Thus, there is

overall significance at  = .01. One of the three predictors is
significant. Gasoline Consumption has a t = 2.67 with p-value of .032
which is statistically significant at  = .05. The p-values of the t
statistics for the other two predictors are insignificant indicating that a
model with just

Gasoline Consumption as a single predictor might be nearly as strong.

 P a g e | 492

15.10 The regression model is:

Insider Ownership =

17.68 - 0.0594 Debt Ratio - 0.118 Dividend Payout

The overall value of F is only 0.02 with p-value of .982. This model is
not significant. Neither of the t values are significant (tDebt = -0.19 with
a p-value of .855 and tDividend = -0.11 with a p-value of .913).

15.11 The regression model is:

ŷ
= 3.981 + 0.07322 x1 - 0.03232 x2 - 0.003886 x3

The overall F for this model is 100.47 with is significant at  = .001.

Only one of the predictors, x1, has a significant t value (t = 3.50, p-
value of .005). The other independent variables have non significant t
values

(x2: t = -1.55, p-value of .15 and x3: t = -1.01, p-value of .332). Since
x2 and x3 are non significant predictors, the researcher should consider
the using a simple regression model with only x1 as a predictor. The R2
would drop some but the model would be much more parsimonious.

15.12 The regression equation for the model using both x1 and x2 is:
 P a g e | 493

ŷ
= 243.44 - 16.608 x1 - 0.0732 x2

The overall F = 156.89 with a p-value of .000. x1 is a significant

predictor of y as indicated by t = - 16.10 and a p-value of .000.

For x2, t = -0.39 with a p-value of .702. x2 is not a significant

predictor of y when included with x1. Since x2 is not a significant
predictor, the researcher might want to rerun the model using justx1 as
a predictor.

The regression model using only x1 as a predictor is:

ˆ
Yˆ
= 235.143 - 16.7678 x1

There is very little change in the coefficient of x1 from model one

(2 predictors) to this model. The overall F = 335.47 with a p-value

of .000 is highly significant. By using the one-predictor model, we get
virtually the same predictability as with the two predictor model and it
is more parsimonious.

15.13 There are 3 predictors in this model and 15 observations.

The regression equation is:

ŷ
= 657.053 + 5.7103 x1 – 0.4169 x2 –3.4715 x3

F = 8.96 with a p-value of .0027

 P a g e | 494

x1 is significant at  = .01 (t = 3.19, p-value of .0087)

x3 is significant at  = .05 (t = - 2.41, p-value of .0349)

The model is significant overall.

15.14 The standard error of the estimate is 3.503. R2 is .408 and the
adjusted R2 is only .203. This indicates that there are a lot of
insignificant predictors in the model. That is underscored by the fact
that eight of the nine predictors have non significant t values.

15.15 se = 9.722, R2 = .515 but the adjusted R2 is only .404. The difference
in the two is due to the fact that two of the three predictors in the
model are non-significant. The model fits the data only modestly. The
adjusted R2 indicates that 40.4% of the variance of y is accounted for
by this model and 59.6% is unaccounted for by the model.

15.16 The standard error of the estimate of 14,660.57 indicates that this
model predicts Per Capita Personal Consumption to within + 14,660.57
about 68% of the time. The entire range of Personal Per Capita for the
data is slightly less than 110,000. Relative to this range, the standard
error of the estimate is modest. R2 = .85988 and the adjusted value of
R2 is .799828 indicating that there are potentially some non significant
variables in the model. An examination of the t statistics reveals that
two of the three predictors are not significant. The model has
relatively good predictability.
 P a g e | 495

15.17 se = 6.544. R2 = .005. This model has no predictability.

 P a g e | 496

15.18 The value of se = 0.2331, R2 = .965, and adjusted R2 = .955. This is a

very strong regression model. However, since x2 and x3 are not

significant predictors, the researcher should consider the using a
simple regression model with only x1 as a predictor. The R2 would drop
some but the model would be much more parsimonious.

15.19 For the regression equation for the model using both x1 and x2, se =
6.333,

R2 = .963 and adjusted R2 = .957. Overall, this is a very strong model.

For the regression model using only x1 as a predictor, the standard
error of the estimate is 6.124, R2 = .963 and the adjusted R2 = .960.
The value of R2 is the same as it was with the two predictors. However,
the adjusted R2 is slightly higher with the one-predictor model because
the non-significant variable has been removed. In conclusion, by using
the one predictor model, we get virtually the same predictability as
with the two predictor model and it is more parsimonious.

15.20 R2 = .710, adjusted R2 = .630, se = 109.43. The model is significant

overall. The R2 is higher but the adjusted R2 by 8%. The model is

moderately strong.
 P a g e | 497

15.21 The Histogram indicates that there may be some problem with the
error

terms being normally distributed. The Residuals vs. Fits plot reveals
that there may be some lack of homogeneity of error variance.

15.22 There are four predictors. The equation of the regression model is:

ŷ
= -55.9 + 0.0105 x1 – 0.107 x2 + 0.579 x3 – 0.870 x4

The test for overall significance yields an F = 55.52 with a p-value of .

000

which is significant at  = .001. Three of the t tests for regression

coefficients are significant at  = .01 including the coefficients for

x2, x3, and x4. The R2 value of 80.2% indicates strong predictability for
the model. The value of the adjusted R2 (78.8%) is close to R2 and se is
9.025.

15.23 There are two predictors in this model. The equation of the regression
model is:

ŷ
= 203.3937 + 1.1151 x1 – 2.2115 x2

The F test for overall significance yields a value of 24.55 with an

associated p-value of .0000013 which is significant at  = .00001.

Both
 P a g e | 498

variables yield t values that are significant at a 5% level of

significance.

x2 is significant at  = .001. The R2 is a rather modest 66.3% and the

standard error of the estimate is 51.761.

15.24 The regression model is:

ŷ
= 137.27 + 0.0025 x1 + 29.206 x2

F = 10.89 with p = .005, se = 9.401, R2 = .731, adjusted R2 = .664. For

x1, t = 0.01 with p = .99 and for x2, t = 4.47 with p = .002. This model
has good predictability. The gap between R2 and adjusted R2 indicates
that there may be a non-significant predictor in the model. The t
values show x1 has virtually no predictability and x2 is a significant
predictor of y.

15.25 The regression model is:

Yˆ
= 362.3054 – 4.745518 x1 - 13.89972 x2 + 1.874297 x3
 P a g e | 499

F = 16.05 with p = .001, se = 37.07, R2 = .858, adjusted R2 = .804. For

x1, t = -4.35 with p = .002; for x2, t = -0.73 with p = .483, for x3, t =
1.96 with p = .086. Thus, only one of the three predictors, x1, is a
significant predictor in this model. This model has very good
predictability (R2 = .858). The gap between R2 and adjusted R2
underscores the fact that there are two non-significant predictors in
this model.
 P a g e | 500

15.26 The overall F for this model was 12.19 with a p-value of .002 which is
significant at  = .01. The t test for Silver is significant at  = .01 ( t =
4.94, p = .001). The t test for Aluminum yields a t = 3.03 with a p-
value of .016 which is significant at  = .05. The t test for Copper was
insignificant with a p-value of .939. The value of R2 was 82.1%
compared

to an adjusted R2 of 75.3%. The gap between the two indicates the

presence of some insignificant predictors (Copper). The standard error
of the estimate is 53.44.

15.27 The regression model was:

Employment = 71.03 + 0.4620 NavalVessels + 0.02082 Commercial

F = 1.22 with p = .386 (not significant)

R2 = .379 and adjusted R2 = .068

The low value of adjusted R2 indicates that the model has very low
predictability. Both t values are not significant (tNavalVessels = 0.67 with

p = .541 and tCommercial = 1.07 with p = .345). Neither predictor is a

significant predictor of employment.
 P a g e | 501

15.28 The regression model was:

All = -1.06 + 0.475 Food + 0.250 Shelter – 0.008 Apparel +

0.272 Fuel Oil

F = 97.98 with a p-value of .000

se = 0.7472, R2 = .963 and adjusted R2 = .953

One of the predictor variables, Food, produces a t value that is

significant at  = .001. Two others are significant at  = .05: Shelter (t
= 2.48, p-value of .025 and Fuel Oil (t = 2.36 with a p-value of .032).
 P a g e | 502

15.29 The regression model was:

Corn = -2718 + 6.26 Soybeans – 0.77 Wheat

F = 14.25 with a p-value of .003 which is significant at  = .01

se = 862.4, R2 = 80.3%, adjusted R2 = 74.6%

One of the two predictors, Soybeans, yielded a t value that was

significant at  = .01 while the other predictor, Wheat was not
significant (t = -0.75 with a p-value of .476).

15.30 The regression model was:

Grocery = 76.23 + 0.08592 Housing + 0.16767 Utility

+ 0.0284 Transportation - 0.0659 Healthcare

F = 2.29 with p = .095 which is not significant at  = .05.

se = 4.416, R2 = .315, and adjusted R2 = .177.

Only one of the four predictors has a significant t ratio and that is
Utility with t = 2.57 and p = .018. The ratios and their respective
probabilities are:

thousing = 1.68 with p = .109, ttransportation = 0.17 with p = .87, and

 P a g e | 503

thealthcare = - 0.64 with p = .53.

This model is very weak. Only the predictor, Utility, shows much
promise in accounting for the grocery variability.

15.31 The regression equation is:

ŷ
= 87.89 – 0.256 x1 – 2.714 x2 + 0.0706 x3

F = 47.57 with a p-value of .000 significant at  = .001.

se = 0.8503, R2 = .941, adjusted R2 = .921.

All three predictors produced significant t tests with two of them

(x2 and x3) significant at .01 and the other, x1 significant at  = .05.
This is

a very strong model.

15.32 Two of the diagnostic charts indicate that there may be a problem
with the
error terms being normally distributed. The histogram indicates that
the error term distribution might be skewed to the right and the normal
probability plot is somewhat nonlinear. In addition, the residuals vs.
fits chart indicates a potential heteroscadasticity problem with
 P a g e | 504

residuals for middle values of x producing more variability that those

for lower and higher values of x.

Chapter 16
Building Multiple Regression Models

LEARNING OBJECTIVES

This chapter presents several advanced topics in multiple regression analysis

enabling you to:

1. Analyze and interpret nonlinear variables in multiple regression

analysis.

2. Understand the role of qualitative variables and how to use them in

multiple regression analysis.

3. Learn how to build and evaluate multiple regression models.

4. Detect influential observations in regression analysis.

 P a g e | 505

CHAPTER TEACHING STRATEGY

In chapter 15, the groundwork was prepared for chapter 16 by

presenting multiple regression models along with mechanisms for testing the
strength of the models such as se, R2, t tests of the regression coefficients,
and the residuals.

The early portion of this chapter is devoted to nonlinear regression

models and search procedures. There are other exotic types of regression
models that can be explored. It is hoped that by studying section 16.1, the
student will be somewhat prepared to explore other nonlinear models on
his/her own. Tukey’s Ladder of Transformations can be useful in steering the
research towards particular recoding schemes that will result in better fits for
the data.

Dummy or indicator variables can be useful in multiple regression

analysis. Remember to emphasize that only one dummy variable is used to
represent two categories (yes/no, male/female, union/nonunion, etc.). For c
categories of a qualitative variable, only c-1 indicator variables should be
included in the multiple regression model.

Several search procedures have been discussed in the chapter

including stepwise regression, forward selection, backward elimination, and
all possible regressions. All possible regressions is presented mainly to
demonstrate to the student the large number of possible models that can be
examined. Most of the effort is spent on stepwise regression because of its
common usage. Forward selection is presented as the same as stepwise
regression except that forward selection procedures do not go back and
examine variables that have been in the model at each new step. That is,
with forward selection, once a variable is in the model, it stays in the model.
Backward elimination begins with a "full" model of all predictors. Sometimes
there may not be enough observations to justify such a model.
 P a g e | 506

CHAPTER OUTLINE

16.1 Non Linear Models: Mathematical Transformation

Polynomial Regression

Tukey’s Ladder of Transformations

Regression Models with Interaction

Model Transformation

16.2 Indicator (Dummy) Variables

16.3 Model-Building: Search Procedures

Search Procedures

All Possible Regressions

Stepwise Regression

Forward Selection

Backward Elimination

16.4 Multicollinearity
 P a g e | 507

KEY TERMS

All Possible Regressions Qualitative Variable

Backward Elimination
Search Procedures
Dummy Variable Stepwise Regression

Forward Selection Tukey’s Four-quadrant Approach

Indicator Variable Tukey’s Ladder of

Multicollinearity Transformations

Quadratic Model Variance Inflation Factor

SOLUTIONS TO PROBLEMS IN CHAPTER 16

16.1 Simple Regression Model:

ŷ
= - 147.27 + 27.128 x

F = 229.67 with p = .000, se = 27.27, R2 = .97, adjusted R2 = .966, and

 P a g e | 508

t = 15.15 with p = .000. This is a very strong simple regression model.

Quadratic Model (Using both x and x2):

ŷ
= - 22.01 + 3.385 X + 0.9373 x2

F = 578.76 with p = .000, se = 12.3, R2 = .995, adjusted R2 = .993, for

x:

t = 0.75 with p = .483, and for x2: t = 5.33 with p = .002. The
quadratic model is also very strong with an even higher R2 value.
However, in this model only the x2 term is a significant predictor.
 P a g e | 509

16.2 The model is:

ŷ
= b0b1x

Using logs: log y = log b0 + x log b1

The regression model is solved for in the computer using the values of
x and the values of log y. The resulting regression equation is:

log y = 0.5797 + 0.82096 x

F = 68.83 with p = .000, se = 0.1261, R2 = .852, and adjusted R2 = .

839. This model has relatively strong predictability.
 P a g e | 510

16.3 Simple regression model:

ˆ
Yˆ
= - 1456.6 + 71.017 x

R2 = .928 and adjusted R2 = .910. t = 7.17 with p = .002.

Quadratic regression model:

ŷ
= 1012 - 14.06 x + 0.6115 x2

R2 = .947 but adjusted R2 = .911. The t ratio for the x term is t = -

0.17 with p = .876. The t ratio for the x2 term is t = 1.03 with p = .
377

Neither predictor is significant in the quadratic model. Also, the

adjusted R2 for this model is virtually identical to the simple regression
model. The quadratic model adds virtually no predictability that the
simple regression model does not already have. The scatter plot of the
data follows:
 P a g e | 511

7000

6000

5000
Ad Exp

4000

3000

2000

1000
30 40 50 60 70 80 90 100 110
Eq & Sup Exp

16.4 The model is:

ŷ
= b0b1x

Using logs: log y= log b0 + x log b1

 P a g e | 512

The regression model is solved for in the computer using the values of
x and the values of log y where x is failures and y is liabilities. The
resulting regression equation is:

log liabilities = 3.1256 + 0.012846 failures

F = 19.98 with p = .001, se = 0.2862, R2 = .666, and adjusted R2 = .

633. This model has modest predictability.
 P a g e | 513

16.5 The regression model is:

ŷ
= - 28.61 - 2.68 x1 + 18.25 x2 - 0.2135 x12 - 1.533 x22 + 1.226
x1*x2

F = 63.43 with p = .000 significant at  = .001

se = 4.669, R2 = .958, and adjusted R2 = .943

None of the t ratios for this model are significant. They are t(x1) = -
0.25 with p = .805, t(x2) = 0.91 with p = .378, t(x12) = - 0.33 with .745,

t(x22) = - 0.68 with .506, and t(x1*x2) = 0.52 with p = .613. This model
has a high R2 yet none of the predictors are individually significant.

The same thing occurs when the interaction term is not in the model.
None of the t tests are significant. The R2 remains high at .957
indicating

that the loss of the interaction term was insignificant.

16.6 The F value shows very strong overall significance with a p-value of .
00000073. This is reinforced by the high R2 of .910 and adjusted R2 of .
878. An examination of the t values reveals that only one of the
regression coefficients is significant at  = 05 and that is the
interaction term with a p-value of .039. Thus, this model with both
variables, the square of both variables, and the interaction term
contains only one significant t test and that is for interaction.

Without interaction, the R2 drops to .877 and adjusted R2 to .844. With

the interaction term removed, both variable x2 and x22 are significant at
 P a g e | 514

 = .01.

16.7 The regression equation is:

ŷ
= 13.619 - 0.01201 x1 + 2.998 x2

The overall F = 8.43 is significant at  = .01 (p = .009).

se = 1.245, R2 = .652, adjusted R2 = .575

The t ratio for the x1 variable is only t = -0.14 with p = .893. However
the t ratio for the dummy variable, x2 is t = 3.88 with p = .004. The
indicator variable is the significant predictor in this regression model
that has some predictability (adjusted R2 = .575).

16.8 The indicator variable has c = 4 categories as shown by the c-1

= 3 categories of the predictors (x2, x3, x4).

The regression equation is:

ŷ
= 7.909 + 0.581 x1 + 1.458 x2 - 5.881 x3 - 4.108 x4

Overall F = 13.54, p = .000 significant at  = .001

 P a g e | 515

se = 1.733, R2 = .806, and adjusted R2 = .747

For the predictors, t = 0.56 with p = .585 for the x1 variable (not
significant), t = 1.32 with p = .208 for the first indicator variable (x2)
and is non significant, t = -5.32 with p = .000 for x3 the second
indicator variable and this is significant at  = .001, t = -3.21 with p = .
007 for the third indicator variable (x4) which is significant at  = .01.
This model has strong predictability and the only significant predictor
variables are the two dummy variables, x3 and x4.

16.9 This regression model has relatively strong predictability as indicated

by R2 = .795. Of the three predictor variables, only x1 and x2 have
significant t ratios (using  = .05). x3 (a non indicator variable) is not a
significant predictor. x1, the indicator variable, plays a significant role
in this model along with x2.
 P a g e | 516

16.10 The regression model is:

ŷ
= 41.225 + 1.081 x1 – 18.404 x2

F = 8.23 with p = .0017 which is significant at  = .01. se = 11.744,

R2 = .388 and the adjusted R2 = .341.

The t-ratio for x2 (the dummy variable) is –4.05 which has an

associated

p-value of .0004 and is significant at  = .001. The t-ratio of 0.80 for x1

is not significant (p-value = .4316). With x2 = 0, the regression model

ŷ
becomes = 41.225 + 1.081x1. With x2 = 1, the regression model

ŷ
becomes = 22.821 + 1.081x1. The presence of x2 causes the y

intercept to drop by 18.404. The graph of each of these models

(without

the dummy variable and with the dummy variable equal to one) is
shown

below:
 P a g e | 517

16.11 The regression equation is:

Price = 7.066 - 0.0855 Hours + 9.614 ProbSeat + 10.507 FQ

The overall F = 6.80 with p = .009 which is significant at  = .01. se =

4.02, R2 = .671, and adjusted R2 = .573. The difference between R2
and adjusted R2 indicates that there are some non-significant
predictors in the model. The t ratios, t = - 0.56 with p = .587 and t =
1.37 with p = .202, of Hours and Probability of Being Seated are non-
significant at  = .05. The only significant predictor is the dummy
variable, French Quarter or not, which has a t ratio of 3.97 with p = .
003 which is significant at  = .01. The positive coefficient on this
variable indicates that being in the French Quarter adds to the price of
a meal.

16.12 There will be six predictor variables in the regression analysis:

three for occupation, two for industry, and one for marital status. The
dependent variable is job satisfaction. In total, there will be seven
variables in this analysis.
 P a g e | 518

16.13 Stepwise Regression:

Step 1: x2 enters the model, t = - 7.35 and R2 = .794

The model is = 36.15 - 0.146 x2

Step 2: x3 enters the model and x2 remains in the model.

t for x2 is -4.60, t for x3 is 2.93. R2 = .876.

ŷ
The model is = 26.40 - 0.101 x2 + 0.116 x3

The variable, x1, never enters the procedure.

16.14 Stepwise Regression:

Step 1: x4 enters the model, t = - 4.20 and R2 = .525

ŷ
The model is = 133.53 - 0.78 x4

Step 2: x2 enters the model and x4 remains in the model.

t for x4 is - 3.22 and t for x2 is 2.15. R2 = .637

ŷ
The model is = 91.01 - 0.60 x4 + 0.51 x2
 P a g e | 519

The variables, x1 and x3 never enter the procedure.

16.15 The output shows that the final model had four predictor variables, x4,
x2, x5, and x7. The variables, x3 and x6 did not enter the stepwise
analysis. The procedure took four steps. The final model was:

y1 = - 5.00 x4 + 3.22 x2 + 1.78 x5 + 1.56 x7

The R2 for this model was .5929, and se was 3.36. The t ratios were:

tx4 = 3.07, tx2 = 2.05, tx5 = 2.02, and tx7 = 1.98.

 P a g e | 520

16.16 The output indicates that the stepwise process only went two steps.
Variable x3 entered at step one. However, at step two, x3 dropped out
of the analysis and x2 and x4 entered as the predictors. x1 was the
dependent variable. x5 never entered the procedure and was not
included in the final model as x3 was not. The final regression model
was:

Yˆ
= 22.30 + 12.38 x2 + 0.0047 x4.

R2 = .682 and se = 9.47. tx2 = 2.64 and tx4 = 2.01.

16.17 The output indicates that the procedure went through two steps. At
step 1, dividends entered the process yielding an r2 of .833 by itself.
ŷ
The t value was 6.69 and the model was = - 11.062 + 61.1 x1. At
step 2, net income entered the procedure and dividends remained in
the model. The R2 for this two-predictor model was .897 which is a
modest increase from the simple regression model shown in step one.
The step 2 model was:

Premiums earned = - 3.726 + 45.2 dividends + 3.6 net income

For step 2, tdividends = 4.36 (p-value = .002)

and tnet income = 2.24 (p-value = .056).

correlation matrix

Premiums Income Dividends Gain/Loss

Premiums 1

Income 0.808236 1
 P a g e | 521

Dividends 0.912515 0.682321 1

Gain/Loss -0.40984 0.0924 -0.52241 1

16.18 This stepwise regression procedure only went one step. The only
significant predictor was natural gas. No other predictors entered the
model. The regression model is:

Electricity = 1.748 + 0.994 Natural Gas

For this model, R2 = .9295 and se = 0.490. The t value for natural gas
was 11.48.
 P a g e | 522

Chapter 18
Statistical Quality Control

LEARNING OBJECTIVES

Chapter 18 presents basic concepts in quality control, with a particular

emphasis on statistical quality control techniques, thereby enabling you to:

1. Understand the concepts of quality, quality control, and total quality

management.

2. Understand the importance of statistical quality control in total quality

management.

3. Learn about process analysis and some process analysis tools.

x
4. Learn how to construct charts, R charts, p charts, and c charts.

5. Understand the theory and application of acceptance sampling.

CHAPTER TEACHING STRATEGY

 P a g e | 523

The objective of this chapter is to present the major concepts of

statistical quality control including control charts and acceptance sampling in
a context of total quality management. Too many texts focus only on a few
statistical quality control procedures and fail to provide the student with a
managerial, decision-making context within which to use quality control
statistical techniques. In this text, the concepts of total quality management
along with some varying definitions of quality and some of the major theories
in quality are presented. From this, the student can formulate a backdrop for
the statistical techniques presented. Some statistics’ instructors argue that
students are exposed to some of this material in other courses. However, the
background material on quality control is relatively brief; and at the very
least, students should be required to read over these pages before beginning
the study of control charts.

The background material helps the students understand that everyone

does not agree on what a quality product is. After all, if there is no
agreement on what is quality, then it is very difficult to ascertain or measure
if it is being accomplished. The notion of in-process quality control helps the
student understand why we generate the data that we use to construct
control charts. Once the student is in the work world, it will be incumbent
upon him/her to determine what measurements should be taken and
monitored. A discussion on what types of measurements can be garnered in
a particular business setting might be worthy of some class time. For
example, if a hospital lab wants to improve quality, how would they go about
it? What measurements might be useful? How about a production line of
computer chips?

The chapter contains "some important quality concepts". The attempt

is to familiarize, if only in passing, the student with some of the more well-
known quality concepts. Included in the chapter are such things as team-
building, benchmarking, just-in-time, reengineering, FMEA, Six Sigma, and
Poka-Yoke all of which can effect the types of measurements being taken and
the statistical techniques being used. It is a disservice to send students into
the business world armed with statistical techniques such as acceptance
sampling and control charts but with their heads in the sand about how the
techniques fit into the total quality picture.

Chapter 18 contains a section on process analysis. Improving quality

usually involves an investigation of the process from which the product
emerges. The most obvious example of a process is a manufacturing
 P a g e | 524

assembly line. However, even in most service industries such insurance,

banking, or healthcare there are processes. A useful class activity might be
to brainstorm about what kind of process is involved in a person buying
gasoline for their car, checking in to a hospital, or purchasing a health club
membership. Think about it from a company’s perspective. What activities
must occur in order for a person to get their car filled up?

In analyzing process, we first discuss the construction of flowcharts.

Flowcharting can be very beneficial in identifying activities and flows that
need to be studied for quality improvement. One very important outcome of
a flowchart is the identification of bottlenecks. You may find out that all
applications for employment, for example, must pass across a clerk’s desk
where they sit for several days. This backs up the system and prevents flow.
Other process techniques include fishbone diagrams, Pareto analysis, and
control charts.

In this chapter, four types of control charts are presented. Two of the
charts, the x bar chart and the R chart, deal with measurements of product
attributes such as weight, length, temperature and others. The other two
charts deal with whether or not items are in compliance with specifications (p
chart) or the number of noncompliances per item (c chart). The c chart is
less widely known and used than the other three. As part of the material on
control charts, a discussion on variation is presented. Variation is one of the
main concerns of quality control. A discussion on various types of variation
that can occur in a business setting can be profitable in helping the student
understand why particular measurements are charted and controlled.

CHAPTER OUTLINE

18.1 Introduction to Quality Control

What is Quality Control?

Total Quality Management

Some Important Quality Concepts

 P a g e | 525

Benchmarking

Just-in-Time Systems

Reengineering

Failure Mode and Effects Analysis (FMEA)

Poka-Yoke

Six Sigma

Design for Six Sigma

Lean Manufacturing

Team Building

18.2 Process Analysis

Flowcharts

Pareto Analysis

Cause-and-Effect (Fishbone) Diagrams

Control Charts

Check Sheets

Histogram

Scatter Chart

18.3 Control Charts

Variation

Types of Control Charts

x
Chart

R Charts

p Charts
 P a g e | 526

c Charts

Interpreting Control Charts

18.4 Acceptance Sampling

Single Sample Plan

Double-Sample Plan

Multiple-Sample Plan

Determining Error and OC Curves

 P a g e | 527

KEY TERMS

Acceptance Sampling p Chart

After-Process Quality Control Pareto Analysis

Benchmarking Pareto Chart

c Chart Poka-Yoke

Cause-and-Effect Diagram Process

Centerline Producer’s Risk

Check Sheet Product Quality

Consumer's Risk Quality

Control Chart Quality Circle

Design for Six Sigma Quality Control

Double-Sample Plan R Chart

Failure Mode and Effects Analysis Reengineering

Fishbone Diagram Scatter Chart

Flowchart Single-Sample Plan

Histogram Six Sigma

In-Process Quality Control Team Building

Ishikawa Diagram Total Quality Management

Just-in-Time Inventory Systems Transcendent Quality

Lean Manufacturing Upper Control Limit (UCL)

Lower Control Limit (LCL) User Quality

Manufacturing Quality Value Quality

x
Multiple-Sample Plan Chart
 P a g e | 528

Operating Characteristic (OC) Curve

 P a g e | 529

SOLUTIONS TO PROBLEMS IN CHAPTER 18

18.2 Complaint Number % of Total

Busy Signal 420 56.45

Too long a Wait 184 24.73

Could not get through 85 11.42

Get Disconnected 37 4.97

Transferred to the Wrong Person 10 1.34

Poor Connection 8 1.08

Total 744 99.99

 P a g e | 530

x1 x2 x3 x4 x5
18.4 = 27.00, = 24.29, = 25.29, = 27.71, = 25.86

R1 = 8, R2 = 8, R3 = 9, R4 = 7, R5 = 6

x R
= 26.03 = 7.6

x
For Chart: Since n = 7, A2 = 0.419

x
Centerline: = 26.03

x R
UCL: + A2 = 26.03 + (0.419)(7.6) = 29.21

x R
LCL: - A2 = 26.03 – (0.419)(7.6) = 22.85

For R Chart: Since n = 7, D3 = 0.076 D4 = 1.924

R
Centerline: = 7.6
 P a g e | 531

R
UCL: D4 = (1.924)(7.6) = 14.62

R
LCL: D3 = (0.076)(7.6) = 0.58

x
Chart:

R Chart:
P a g e | 532
 P a g e | 533

x1 x2 x3 x4
18.5 = 4.55, = 4.10, = 4.80, = 4.70,

x5 x6 x7
= 4.30, = 4.73, = 4.38

R1 = 1.3, R2 = 1.0, R3 = 1.3, R4 = 0.2, R5 = 1.1, R6 = 0.8, R7 = 0.6

x R
= 4.51 = 0.90

x
For Chart: Since n = 4, A2 = 0.729

x
Centerline: = 4.51

x R
UCL: + A2 = 4.51 + (0.729)(0.90) = 5.17

x R
LCL: - A2 = 4.51 – (0.729)(0.90) = 3.85

For R Chart: Since n = 4, D3 = 0 D4 = 2.282

R
Centerline: = 0.90
 P a g e | 534

R
UCL: D4 = (2.282)(0.90) = 2.05

R
LCL: D3 = 0

x
Chart:
 P a g e | 535

R Chart:

p̂1 p̂2 p̂3 p̂4 p̂5

18.6 = .02, = .07, = .04, = .03, = .03

p̂6 p̂7 p̂8 p̂9 p̂10

= .05, = .02, = .00, = .01, = .06

p = .033

Centerline: p = .033

(.033)(. 967)
100
UCL: .033 + 3 = .033 + .054 = .087
 P a g e | 536

(.033)(.967)
100
LCL: .033 - 3 = .033 - .054 = .000

pChart:
 P a g e | 537

p̂1 p̂2 p̂3 p̂4

18.7 = .025, = .000, = .025, = .075,

p̂5 p̂6 p̂7

= .05, = .125, = .05

p = .050

Centerline: p = .050

(.05)(.95)
40
UCL: .05 + 3 = .05 + .1034 = .1534

(.05)(. 95)
40
LCL: .05 - 3 = .05 - .1034 = .000

p Chart:
 P a g e | 538

22
c 35
18.8 = = 0.62857

c
Centerline: = 0.62857

c3 c 0.62857
UCL: = 0.62857 + 3 =

0.62857 + 2.37847 = 3.00704

c3 c 0.62857
LCL: = 0.62857 - 3 =

0.62857 – 2.37847 = .000

c Chart:
 P a g e | 539

43
c 32
18.9 = = 1.34375

c
Centerline: = 1.34375

c3 c 1.34375
UCL: = 1.34375 + 3 =

1.34375 + 3.47761 = 4.82136

c3 c 1.34375
LCL: = 1.34375 - 3 =

1.34375 - 3.47761 = 0.000

 P a g e | 540

c Chart:

18.10 a.) Six or more consecutive points are decreasing. Two of three

consecutive points are in the outer one-third (near LCL). Four out of

five points are in the outer two-thirds (near LCL).

b.) This is a relatively healthy control chart with no obvious rule

violations.

c.) One point is above the UCL. Two out of three consecutive points are
in

the outer one-third (both near LCL and near UCL). There are six

consecutive increasing points.

18.11 While there are no points outside the limits, the first chart exhibits
some

problems. The chart ends with 9 consecutive points below the

centerline.
 P a g e | 541

Of these 9 consecutive points, there are at least 4 out of 5 in the outer

2/3 of the lower region. The second control chart contains no points
outside the control limit. However, near the end, there are 8
consecutive points above the centerline. The p chart contains no
points outside the upper control limit. Three times, the chart contains
two out of three points in the outer third. However, this occurs in the
lower third where the proportion of noncompliance items approaches
zero and is probably not a problem to be concerned about. Overall,
this seems to display a process that is in control. One concern might
be the wide swings in the proportions at samples 15, 16 and 22 and
23.

18.12 For the first sample:

If x1 > 4 then reject

If x1 < 2 then accept

If 2 < x1 < 4 then take a second sample

For the second sample, c2 = 3:

If x1 + x2 < c2 then accept

If x1 + x2 > c2 then reject

 P a g e | 542

But x1 = 2 and x2 = 2 so x1 + x2 = 4 > 3

Reject the lot because x1 + x2 = 4 > c2 = 3

This is a double sample acceptance plan

18.13 n = 10 c=0 p0 = .05

P(x = 0) = C0(.05)0(.95)10 = .5987

10

1 - P(x = 0) = 1 - .5987 = .4013

The producer's risk is .4013

p1 = .14

P(x = 0) = C0(.14)0(.86)10 =
15 .2213

The consumer's risk is .2213

18.14 n = 12 c=1 p0 = .04

Producer's Risk = 1 - [P(x = 0) + P(x = 1)] =

 P a g e | 543

1 - [12C0(.04)0(.96)12 + 12 C1(.04)1(.96)11] =

1 - [.6127 + .30635] = 1 - .91905 = .08095

p1 = .15

Consumer's Risk = P(x = 0) + P(x = 1) = C0(.15)0(.85)12 +

12

C1(.15)1(.85)11 = .14224 + .30122 = .44346

12
 P a g e | 544

18.15 n = 8 c=0 p0 = .03 p1 = .1

p Probability

.01 .9227

.02 .8506

.03 .7837

.04 .7214 Producer's Risk for (p0 = .03) =

.05 .6634 1 - .7837
= .2163

.06 .6096

.07 .5596

.08 .5132 Consumer's Risk for (p1 = .10) = .4305

.09 .4703

.10 .4305
.11 .3937

.12 .3596

.13 .3282

.14 .2992

.15 .2725

OC Chart:
P a g e | 545
 P a g e | 546

18.16 n = 11 c=1 p0 = .08 p1 = .20

p Probability

.02 .9805

.04 .9308

.06 .8618

.08 .7819 Producer's Risk for (p0 = .08) = 1 - .7819 = .

2181

.10 .6974

.12 .6127

.14 .5311

.16 .4547 Consumer's Risk for (p1 = 20) = .3221

.18 .3849

.20 .3221

.22 .2667
.24 .2186

OC Chart:
P a g e | 547
 P a g e | 548

18.17

Stop

(no)

D  K  L  M (yes)  Stop

 Stop

 

(no) (no)

Start  A  B (yes) C  E  F  G

(yes)

H(no) J  Stop

(yes) 

 

I     
 P a g e | 549

18.18 Problem Frequency Percent of Total

1 673 26.96

2 29 1.16

3 108 4.33

4 379 15.18

5 73 2.92

6 564 22.60

7 12 0.48

8 402 16.11

9 54 2.16

10 202 8.09

2496
 P a g e | 550

Pareto Chart:

18.19 Fishbone Diagram:

P a g e | 551
 P a g e | 552

18.20 a) n = 13, c = 1, p0 = .05, p1 = .12

Under p0 = .05, the probability of acceptance is:

13 C0(.05)0(.95)13 + 13 C1(.05)1(.95)12 =

(1)(1)(.51334) + (13)(.05)(.54036) =

.51334 + .35123 = .86457

The probability of being rejected = 1 - .86457 = .13543

Under p1 = .12, the probability of acceptance is:

13C0(.12)0(.88)13 + C1(.12)1(.88)12 =
13

(1)(1)(.18979) + (13)(.12)(.21567) =

.18979 + .33645 = .52624

The probability of being rejected = 1 - .52624 = .47376

b) n = 20, c = 2, p0 = .03

The probability of acceptance is:

20 C0(.03)0(.97)19 + 20 C1(.03)1(.97)19 + C2(.03)2(.97)18 =

20

(1)(1)(.54379) + (20)(.03)(.56061) + (190)(.0009)(.57795) =

.54379 + .33637 + .09883 = .97899

 P a g e | 553

The probability of being rejected = 1 - .97899 = .02101

which is the producer’s risk

p̂1 p̂2 p̂3 p̂4 p̂5

18.21 = .06, = .22, = .14, = .04, = .10,

p̂6 p̂7 p̂8 p̂9 p̂10

= .16, = .00, = .18, = .02, = .12

52
500
p= = .104

Centerline: p = .104

(. 104)(. 896)
50
UCL: .104 + 3 = .104 + .130 = .234

(.104)(.896)
50
LCL: .104 - 3 = .104 - .130 = .000
 P a g e | 554

p Chart:

x1 x2 x3 x4
18.22 = 24.022, = 24.048, = 23.996, = 24.000,

x5 x6 x7 x8
= 23.998, = 24.018, = 24.000, = 24.034,

x9 x 10 x 11 x12
= 24.014, = 24.002, = 24.012, = 24.022

R1 = .06, R2 = .09, R3 = .08, R4 = .03, R5 = .05, R6 = .05,

R7 = .05, R8 = .08, R9 = .03, R10 = .01, R11 = .04, R12 = .05

 P a g e | 555

x R
= 24.01383 = 0.05167

x
For Chart: Since n = 12, A2 = .266

x
Centerline: = 24.01383

x R
UCL: + A2 = 24.01383 + (0.266)(.05167) =

24.01383 + .01374 = 24.02757

x R
LCL: - A2 = 24.01383 - (0.266)(.05167) =

24.01383 - .01374 = 24.00009

For R Chart: Since n = 12, D3 = .284 D4 = 1.716

R
Centerline: = .05167

R
UCL: D4 = (1.716)(.05167) = .08866

R
LCL: D3 = (.284)(.05167) = .01467
 P a g e | 556

x
Chart:

R Chart:
 P a g e | 557

18.23 n = 15, c = 0, p0 = .02, p1 = .10

p Probability

.01 .8601

.02 .7386

.04 .5421

.06 .3953

.08 .2863

.10 .2059

.12 .1470

.14 .1041

Producer's Risk for (p0 = .02) = 1 - .7386 = .2614

Consumer's Risk for (p1 = .10) = .2059

OC Curve:
P a g e | 558
 P a g e | 559

77
c 36
18.24 = = 2.13889

c
Centerline: = 2.13889

c3 c 2.13889
UCL: = 2.13889 + 3 =

2.13889 + 4.38748 = 6.52637

c3 c 2.13889
LCL: = 2.13889 - 3 =

2.13889 – 4.38748 = .00000

c Chart:
 P a g e | 560

x1 x2 x3 x4
18.25 = 1.2100, = 1.2050, = 1.1900, = 1.1725,

x5 x6 x7 x8
= 1.2075, = 1.2025, = 1.1950, = 1.1950,

x9
= 1.1850

R1 = .04, R2 = .02, R3 = .04, R4 = .04, R5 = .06, R6 = .02,

R7 = .07, R8 = .07, R9 = .06,

x R
= 1.19583 = 0.04667

x
For Chart: Since n = 4, A2 = .729

x
Centerline: = 1.19583

x R
UCL: + A2 = 1.19583 + .729(.04667) =

1.19583 + .03402 = 1.22985

 P a g e | 561

x R
LCL: - A2 = 1.19583 - .729(.04667) =

1.19583 - .03402 = 1.16181

For R Chart: Since n = 9, D3 = .184 D4 = 1.816

R
Centerline: = .04667

R
UCL: D4 = (1.816)(.04667) = .08475

R
LCL: D3 = (.184)(.04667) = .00859

x
Chart:
 P a g e | 562

R chart:

x1 x2 x3 x4
18.26 = 14.99333, = 15.00000, = 14.97833, = 14.97833,

x5 x6 x7 x8
= 15.01333, = 15.00000, = 15.01667, = 14.99667,

R1 = .03, R2 = .07, R3 = .05, R4 = .05,

R5 = .04, R6 = .05, R7 = .05, R8 = .06

x R
= 14.99854 = 0.05

x
For Chart: Since n = 6, A2 = .483
 P a g e | 563

x
Centerline: = 14.99854

x R
UCL: + A2 = 14.99854 + .483(.05) =

14.00854 + .02415 = 15.02269

x R
LCL: - A2 = 14.99854 - .483(.05) =

14.00854 - .02415 = 14.97439

For R Chart: Since n = 6, D3 = 0 D4 = 2.004

R
Centerline: = .05

R
UCL: D4 = 2.004(.05) = .1002

R
LCL: D3 = 0(.05) = .0000

x
Chart:
 P a g e | 564

R chart:

p̂1 p̂2 p̂3 p̂4

18.27 = .12, = .04, = .00, = .02667,

p̂5 p̂6 p̂7 p̂8

= .09333, = .18667, = .14667, = .10667,

p̂9 p̂10 p̂11 p̂12

= .06667, = .05333, = .0000, = .09333
 P a g e | 565

70
900
p= = .07778 Centerline: p = .07778

(.07778)(.92222)
75
UCL: .07778 + 3 = .07778 + .09278 = .17056

(.07778)(.92222)
75
LCL: .07778 - 3 = .07778 - .09278 = .00000

p Chart:
 P a g e | 566

16
c 25
18.28 = = 0.64

c
Centerline: = 0.64

c3 c 0.64
UCL: = 0.64 + 3 =

0.64 + 2.4 = 3.04

c3 c 0.64
LCL: = 0.64 - 3 =

0.64 – 2.4 = .00000

c Chart:
P a g e | 567
 P a g e | 568

18.29 n = 10 c=2 p0 = .10 p1 = .30

p Probability

.05 .9885

.10 .9298

.15 .8202

.20 .6778

.25 .5256

.30 .3828

.35 .2616

.40 .1673

.45 .0996

.50 .0547

Producer's Risk for (p0 = .10) = 1 - .9298 = .0702

Consumer's Risk for (p1 = .30) = .3828

P a g e | 569
 P a g e | 570

81
c 40
18.30 = = 2.025

c
Centerline: = 2.025

c3 c 2.025
UCL: = 2.025 + 3 =

2.025 + 4.26907 = 6.29407

c3 c 2.025
LCL: = 2.025 - 3 =

2.025 – 4.26907 = .00000

c Chart:
 P a g e | 571

p̂1 p̂2 p̂3 p̂4

18.31 = .05, = .00, = .15, = .075,

p̂5 p̂6 p̂7 p̂8

= .025, = .025, = .125, = .00,

p̂9 p̂10 p̂11 p̂12

= .10, = .075, = .05, = .05,

p̂13 p̂14 p̂15

= .15, = .025, = .000

36
600
p= = .06

Centerline: p = .06

(.06)(.94)
40
UCL: .06 + 3 = .06 + .11265 = .17265
 P a g e | 572

(.06)(. 94)
40
LCL: .06 - 3 = .06 - .112658 = .00000

p Chart:

18.32 The process appears to be in control. Only 1 sample mean is beyond

the outer limits (97% of the means are within the limits). There are no
more than four means in a row on one side of the centerline. There are
no more than five consecutive points decreasing or three consecutive
points increasing. About 2/3 (67%) of the points are within the inner
x
1/3 of the confidence bands (+ 1 ).
 P a g e | 573

18.33 There are some items to be concerned about with this chart. Only one
sample range is above the upper control limit. However, near the
beginning of the chart there are eight sample ranges in a row below
the centerline. Later in the run, there are nine sample ranges in a row
above the centerline. The quality manager or operator might want to
determine if there is some systematic reason why there is a string of
ranges below the centerline and, perhaps more importantly, why there
are a string of ranges above the centerline.

18.34 This p chart reveals that two of the sixty samples (about 3%) produce
proportions that are too large. Nine of the sixty samples (15%)
 p̂
produce proportions large enough to be greater than 1 above the
centerline. In general, this chart indicates a process that is under
control.

18.35 The centerline of the c chart indicates that the process is averaging
0.74 nonconformances per part. Twenty-five of the fifty sampled items
have zero nonconformances. None of the samples exceed the upper
control limit for nonconformances. However, the upper control limit is
3.321 nonconformances, which in and of itself, may be too many.
Indeed, three of the fifty (6%) samples actually had three
nonconformances. An additional six samples (12%) had two
nonconformances. One matter of concern may be that there is a run of
ten samples in which nine of the samples exceed the centerline
(samples 12 through 21). The question raised by this phenomenon is
 P a g e | 574

whether or not there is a systematic flaw in the process that produces

strings of nonconforming items.

Chapter 19
Decision Analysis

LEARNING OBJECTIVES

Chapter 19 describes how to use decision analysis to improve management

decisions, thereby enabling you to:

1. Learn about decision making under certainty, under uncertainty, and

under risk.

2. Learn several strategies for decision-making under uncertainty,

including expected payoff, expected opportunity loss, maximin,
maximax, and minimax regret.
 P a g e | 575

3. Learn how to construct and analyze decision trees.

4. Understand aspects of utility theory.

5. Learn how to revise probabilities with sample information.

CHAPTER TEACHING STRATEGY

The notion of contemporary decision making is built into the title of the
text as a statement of the importance of recognizing that statistical analysis
is primarily done as a decision-making tool. For the vast majority of students,
statistics take on importance only in as much as they aid decision-makers in
weighing various alternative pathways and helping the manager make the
best possible determination. It has been an underlying theme from chapter 1
that the techniques presented should be considered in a decision-making
context. This chapter focuses on analyzing the decision-making situation and
presents several alternative techniques for analyzing decisions under varying
conditions.

Early in the chapter, the concepts of decision alternatives, the states of

nature, and the payoffs are presented. It is important that decision makers
spend time brainstorming about possible decision alternatives that might be
available to them. Sometimes the best alternatives are not obvious and are
not immediately considered. The international focus on foreign companies
investing in the U.S. presents a scenario in which there are several possible
alternatives available. By using cases such as the Fletcher-Terry case at the
chapter's end, students can practice enumerating possible decision
alternatives.
 P a g e | 576

States of nature are possible environments within which the outcomes

will occur over which we have no control. These include such things as the
economy, the weather, health of the CEO, wildcat strikes, competition,
change in consumer demand, etc. While the text presents problems with
only a few states of nature in order to keep the length of solution reasonable,
students should learn to consider as many states of nature as possible in
decision making. Determining payoffs is relatively difficult but essential in
the analysis of decision alternatives.

Decision-making under uncertainty is the situation in which the

outcomes are not known and there are no probabilities given as to the
likelihood of them occurring. With these techniques, the emphasis is whether
or not the approach is optimistic, pessimistic, or weighted somewhere in
between.

In making decisions under risk, the probabilities of each state of nature

occurring are known or are estimated. Decision trees are introduced as an
alternative mechanism for displaying the problem. The idea of an expected
monetary value is that if this decision process were to continue with the same
parameters for a long time, what would the long-run average outcome be?
Some decisions lend themselves to long-run average analysis such as
gambling outcomes or insurance actuary analysis. Other decisions such as
building a dome stadium downtown or drilling one oil well tend to be more
one time activities and may not lend themselves as nicely to expected value
analysis. It is important that the student understand that expected value
outcomes are long-run averages and probably will not occur in single
instance decisions.

Utility is introduced more as a concept than an analytic technique. The

idea here is to aid the decision-maker in determining if he/she tends to be

more of a risk-taker, an EMV'r, or risk-averse. The answer might be that it
depends on the matter over which the decision is being made. One might be
a risk-taker on attempting to employ a more diverse work force and at the
same time be more risk-averse in investing the company's retirement fund.
 P a g e | 577

CHAPTER OUTLINE

19.1 The Decision Table and Decision Making Under Certainty

Decision Table

Decision-Making Under Certainty

19.2 Decision Making Under Uncertainty

Maximax Criterion

Maximin Criterion

Hurwicz Criterion

Minimax Regret

19.3 Decision Making Under Risk

Decision Trees

Expected Monetary Value (EMV)

Expected Value of Perfect Information

Utility

19.4 Revising Probabilities in Light of Sample Information

Expected Value of Sample Information

 P a g e | 578

KEY TERMS

Decision Alternatives Hurwicz Criterion

Decision Analysis Maximax Criterion

Decision Making Under Certainty Maximin Criterion

Decision Making Under Risk Minimax Regret

Decision Making Under Uncertainty Opportunity Loss Table

Decision Table Payoffs

Decision Trees Payoff Table

EMV'er Risk-Avoider

Expected Monetary Value (EMV) Risk-Taker

Expected Value of Perfect Information States of Nature

Expected Value of Sample Information Utility

 P a g e | 579

SOLUTIONS TO PROBLEMS IN CHAPTER 19

19.1 S1 S2 S3 Max Min

d1 250 175 -25 250 -25

d2 110 100 70 110 70

d3 390 140 -80 390 -80

a.) Max {250, 110, 390} = 390 decision: Select d3

b.) Max {-25, 70, -80} = 70 decision: Select d2

c.) For  = .3

d1: .3(250) + .7(-25) = 57.5

d2: .3(110) + .7(70) = 82

d3: .3(390) + .7(-80) = 61

 P a g e | 580

decision: Select d2

For  = .8

d1: .8(250) + .2(-25) = 195

d2: .8(110) + .2(70) = 102

d3: .8(390) + .2(-80) = 296

decision: Select d3

Comparing the results for the two different values of alpha, with a
more pessimist point-of-view ( = .3), the decision is to select d2 and
the payoff is 82. Selecting by using a more optimistic point-of-view (
= .8) results in choosing d3 with a higher payoff of 296.

d.) The opportunity loss table is:

S1 S2 S3 Max

d1 140 0 95 140
 P a g e | 581

d2 280 75 0 280

d3 0 35 150 150

The minimax regret = min {140, 280, 150} = 140

Decision: Select d1 to minimize the regret.

19.2 S1 S2 S3 S4 Max Min

d1 50 70 120 110 120 50

d2 80 20 75 100 100 20

d3 20 45 30 60 60 20

d4 100 85 -30 -20 100 -30

d5 0 -10 65 80 80 -10

a.) Maximax = Max {120, 100, 60, 100, 80} = 120

Decision: Select d1

b.) Maximin = Max {50, 20, 20, -30, -10} = 50

 P a g e | 582

Decision: Select d1

c.)  = .5

Max {[.5(120)+.5(50)], [.5(100)+.5(20)],

[.5(60)+.5(20)], [.5(100)+.5(-30)], [.5(80)+.5(-10)]}=

Max { 85, 60, 40, 35, 35 } = 85

Decision: Select d1
 P a g e | 583

d.) Opportunity Loss Table:November 8, 1996

S1 S2 S3 S4 Max

d1 50 15 0 0 50

d2 20 65 45 10 65

d3 80 40 90 50 90

d4 0 0 150 130 150

d5 100 95 55 30 100

Min {50, 65, 90, 150, 100} = 50

Decision: Select d1
 P a g e | 584

19.3 R D I Max Min

A 60 15 -25 60 -25

B 10 25 30 30 10

C -10 40 15 40 -10

D 20 25 5 25 5

Maximax = Max {60, 30, 40, 25} = 60

Decision: Select A

Maximin = Max {-25, 10, -10, 5} = 10

Decision: Select B
 P a g e | 585

19.4 Not Somewhat Very Max Min

None -50 -50 -50 -50 -50

Few -200 300 400 400 -200

Many -600 100 1000 1000 -600

a.) For Hurwicz criterion using  = .6:

Max {[.6(-50) + .4(-50)], [.6(400) + .4(-200)],

[.6(1000) + .4(-600)]} = {-50, -160, 360}= 360

Decision: Select "Many"

b.) Opportunity Loss Table:

Not Somewhat Very Max

None 0 350 1050 1050

Few 150 0 600 600

Many 550 200 0 550

 P a g e | 586

Minimax regret = Min {1050, 600, 550} = 550

Decision: Select "Many"

 P a g e | 587

19.5, 19.6
 P a g e | 588

19.7 Expected Payoff with Perfect Information =

5(.15) + 50(.25) + 20(.30) + 8(.10) + 6(.20) = 31.75

Expected Value of Perfect Information = 31.25 - 25.25 = 6.50

19.8 a.) & b.)

 P a g e | 589

c.) Expected Payoff with Perfect Information =

150(40) + 450(.35) + 700(.25) = 392.5

Expected Value of Perfect Information = 392.5 - 370 = 22.50

 P a g e | 590

19.9 Down(.30) Up(.65) No Change(.05) EMV

Lock-In -150 200 0 85

No 175 -250 0 -110

Decision: Based on the highest EMV)(85), "Lock-In"

Expected Payoff with Perfect Information =

175(.30) + 200(.65) + 0(.05) = 182.5

Expected Value of Perfect Information = 182.5 - 85 = 97.5

19.10 EMV

No Layoff -960

Layoff 1000 -320

 P a g e | 591

Layoff 5000 400

Decision: Based on maximum EMV (400), Layoff 5000

Expected Payoff with Perfect Information =

100(.10) + 300(.40) + 600(.50) = 430

Expected Value of Perfect Information = 430 - 400 = 30

19.11 a.) EMV = 200,000(.5) + (-50,000)(.5) = 75,000

b.) Risk Avoider because the EMV is more than the

investment (75,000 > 50,000)

c.) You would have to offer more than 75,000 which

is the expected value.

 P a g e | 592

19.12 a.) S1(.30) S2(.70) EMV

d1 350 -100 35

d2 -200 325 167.5

Decision: Based on EMV,

maximum {35, 167.5} = 167.5

b. & c.) For Forecast S1:

Prior Cond. Joint Revised

S1 .30 .90 .27 .6067

S2 .70 .25 .175 .3933

F(S1) = .445

For Forecast S2:

Prior Cond. Joint Revised

 P a g e | 593

S1 .30 .10 .030 .054

S2 .70 .75 .525 .946

F(S2) = .555

EMV with Sample Information = 241.63

 P a g e | 594

d.) Value of Sample Information = 241.63 - 167.5 = 74.13

 P a g e | 595

19.13

Dec(.60) Inc(.40) EMV

S -225 425 35

M 125 -150 15

L 350 -400 50

Decision: Based on EMV = Maximum {35, 15, 50} = 50

For Forecast (Decrease):

Prior Cond. Joint Revised

Decrease .60 .75 .45 .8824

Increase .40 .15 .06 .1176

F(Dec) = .51

For Forecast (Increase):

 P a g e | 596

Prior Cond. Joint Revised

Decrease .60 .25 .15 .3061

Increase .40 .85 .34 .6939

F(Inc) = .49
 P a g e | 597

The expected value with sampling is 244.275

EVSI = EVWS - EMV = 244.275 - 50 = 194.275

P a g e | 598
 P a g e | 599

19.14 Decline(.20) Same(.30) Increase(.50) EMV

Don't Plant 20 0 -40 -16

Small -90 10 175 72.5

Large -600 -150 800 235

Decision: Based on Maximum EMV =

Max {-16, 72.5, 235} = 235, plant a large tree farm

For forecast decrease:

Prior Cond. Joint Revised

.20 .70 .140 .8974

.30 .02 .006 .0385

.50 .02 .010 .0641

P(Fdec) = .156
 P a g e | 600

For forecast same:

Prior Cond. Joint Revised

.20 .25 .05 .1333

.30 .95 .285 .7600

.50 .08 .040 .1067

P(Fsame) = .375

For forecast increase:

Prior Cond. Joint Revised

.20 .05 .01 .0213

.30 .03 .009 .0192

.50 .90 .45 .9595

 P a g e | 601

P(Finc) = .469
 P a g e | 602

The Expected Value with Sampling Information is 360.413

EVSI = EVWSI - EMV = 360.413 - 235 = 125.413

19.15 Oil(.11) No Oil(.89) EMV

Drill 1,000,000 -100,000 21,000

Don't Drill 0 0 0

Decision: The EMV for this problem is Max {21,000, 0} = 21,000.

The decision is to Drill.

Actual

Oil No Oil

Oil .20 .10

Forecast

No Oil .80 .90

 P a g e | 603

Forecast Oil:

State Prior Cond. Joint Revised

Oil .11 .20 .022 .1982

No Oil .89 .10 .089 .8018

P(FOil) = .111

Forecast No Oil:

State Prior Cond. Joint Revised

Oil .11 .80 .088 .0990

No Oil .89 .90 .801 .9010

P(FNo Oil) = .889

 P a g e | 604

The Expected Value With Sampling Information is 21,012.32

EVSI = EVWSI - EMV = 21,000 - 21,012.32 = 12.32

 P a g e | 605

19.16 S1 S2 Max. Min.

d1 50 100 100 50

d2 -75 200 200 -75

d3 25 40 40 25

d4 75 10 75 10

a.) Maximax: Max {100, 200, 40, 75} = 200

Decision: Select d2

b.) Maximin: Max {50, -75, 25, 10} = 50

Decision: Select d1

c.) Hurwicz with  = .6

d1: 100(.6) + 50(.4) = 80

d2: 200(.6) + (-75)(.4) = 90

d3: 40(.6) + 25(.4) = 34

 P a g e | 606

d4: 75(.6) + 10(.4) = 49

Max {80, 90, 34, 49} = 90

Decision: Select d2

d.) Opportunity Loss Table:

S1 S2 Maximum

d1 25 100 100

d2 150 0 150

d3 50 160 160

d4 0 190 190

Min {100, 150, 160, 190} = 100

Decision: Select d1
 P a g e | 607

19.17

b.) d1: 400(.3) + 250(.25) + 300(.2) + 100(.25) = 267.5

d2: 300(.3) + (-100)(.25) + 600(.2) + 200(.25) = 235

Decision: Select d1

c.) Expected Payoff of Perfect Information:

400(.3) + 250(.25) + 600(.2) + 200(.25) = 352.5

 P a g e | 608

Value of Perfect Information = 352.5 - 267.5 = 85

 P a g e | 609

19.18 S1(.40) S2(.60) EMV

d1 200 150 170

d2 -75 450 240

d3 175 125 145

Decision: Based on Maximum EMV =

Max {170, 240, 145} = 240

Select d2

Forecast S1:

State Prior Cond. Joint Revised

S1 .4 .9 .36 .667

S2 .6 .3 .18 .333

P(FS1) = .54

Forecast S2:
 P a g e | 610

State Prior Cond. Joint Revised

S1 .4 .1 .04 .087

S2 .6 .7 .42 .913

P(FS2) = .46
 P a g e | 611

The Expected Value With Sample Information is 285.00

EVSI = EVWSI - EMV = 285 - 240 = 45

P a g e | 612
 P a g e | 613

19.19 Small Moderate Large Min Max

Small 200 250 300 200 300

Modest 100 300 600 100 600

Large -300 400 2000 -300 2000

a.) Maximax: Max {300, 600, 2000} = 2000

Decision: Large Number

Minimax: Max {200, 100, -300} = 200

Decision: Small Number

b.) Opportunity Loss:

Small Moderate Large Max

Small 0 150 1700 1700

Modest 100 100 1400 1400

Large 500 0 0 500

 P a g e | 614

Min {1700, 1400, 500} = 500

Decision: Large Number

c.) Minimax regret criteria leads to the same decision as Maximax.

 P a g e | 615

19.20 No Low Fast Max Min

Low -700 -400 1200 1200 -700

Medium -300 -100 550 550 -300

High 100 125 150 150 100

a.) = .1:

Low: 1200(.1) + (-700)(.9) = -510

Medium: 550(.1) + (-300)(.9) = -215

High: 150(.1) + 100(.9) = 105

Decision: Price High (105)

b.)  = .5:

Low: 1200(.5) + (-700)(.5) = 250

Medium: 550(.5) + (-300)(.5) = 125

High: 150(.5) + 100(.5) = 125

Decision: Price Low (250)

c.)  = .8:
 P a g e | 616

Low: 1200(.8) + (-700)(.2) = 820

Medium: 550(.8) + (-300)(.2) = 380

High: 150(.8) + 100(.2) = 140

Decision: Price Low (820)

d.) Two of the three alpha values (.5 and .8) lead to a decision of pricing
low.

Alpha of .1 suggests pricing high as a strategy. For optimists (high

alphas), pricing low is a better strategy; but for more pessimistic

people,

pricing high may be the best strategy.

 P a g e | 617

19.21 Mild(.75) Severe(.25) EMV

Reg. 2000 -2500 875

Weekend 1200 -200 850

Not Open -300 100 -200

Decision: Based on Max EMV =

Max{875, 850, -200} = 875, open regular hours.

Expected Value with Perfect Information =

 P a g e | 618

2000(.75) + 100(.25) = 1525

Value of Perfect Information = 1525 - 875 = 650

 P a g e | 619

19.22 Weaker(.35) Same(.25) Stronger(.40) EMV

Don't Produce -700 -200 150 -235

Produce 1800 400 -1600

90

Decision: Based on Max EMV = Max {-235, 90} = 90, select Produce.

Expected Payoff With Perfect Information =

1800(.35) + 400(.25) + 150(.40) = 790

Value of Perfect Information = 790 - 90 = 700

P a g e | 620
 P a g e | 621

19.23 Red.(.15) Con.(.35) Inc.(.50) EMV

Automate -40,000 -15,000 60,000 18,750

Do Not 5,000 10,000 -30,000 -10,750

Decision: Based on Max EMV =

Max {18750, -10750} = 18,750, Select Automate

Forecast Reduction:

State Prior Cond. Joint Revised

R .15 .60 .09 .60

C .35 .10 .035 .2333

I .50 .05 .025 .1667

P(F Red) = .150

Forecast Constant:

State Prior Cond. Joint Revised

R .15 .30 .045 .10

C .35 .80 .280 .6222

I .50 .25 .125 .2778

 P a g e | 622

P(F Cons) = .450

Forecast Increase:

State Prior Cond. Joint Revised

R .15 .10 .015 .0375

C .35 .10 .035 .0875

I .50 .70 .350 .8750

P(FInc) = .400
 P a g e | 623

Expected Value With Sample Information = 21,425.55

EVSI = EVWSI - EMV = 21,425.55 - 18,750 = 2,675.55

P a g e | 624
 P a g e | 625

19.24 Chosen(.20) Not Chosen(.80) EMV

Build 12,000 -8,000 -4,000

Don't -1,000 2,000 1,400

Decision: Based on Max EMV = Max {-4000, 1400} = 1,400,

choose "Don't Build" as a strategy.

Forecast Chosen:

State Prior Cond. Joint Revised

Chosen .20 .45 .090 .2195

Not Chosen .80 .40 .320 .7805

P(FC) = .410

Forecast Not Chosen:

State Prior Cond. Joint Revised

Chosen .20 .55 .110 .1864

 P a g e | 626

Not Chosen .80 .60 .480 .8136

P(FC) = .590

Expected Value With Sample Information = 1,400.09

EVSI = EVWSI - EMV = 1,400.09 - 1,400 = .09

P a g e | 627
 P a g e | 628

16.19 y x1 x2 x3

y - -.653 -.891 .821

x1 -.653 - .650 -.615

x2 -.891 .650 - -.688

x3 .821 -.615 -.688 -

There appears to be some correlation between all pairs of the predictor

variables, x1, x2, and x3. All pairwise correlations between independent
variables are in the .600 to .700 range.
 P a g e | 629

16.20 y x1 x2 x3 x4

y - -.241 .621 .278 -.724

x1 -.241 - -.359 -.161 .325

x2 .621 -.359 - .243 -.442

x3 .278 -.161 .243 - -.278

x4 -.724 .325 -.442 -.278 -

An examination of the intercorrelations of the predictor variables

reveals that the highest pairwise correlation exists between variables
x2 and

x4 (-.442). Other correlations between independent variables are less

than .400. Multicollinearity may not be a serious problem in this
regression analysis.
 P a g e | 630

16.21 The stepwise regression analysis of problem 16.17 resulted in two

of the three predictor variables being included in the model. The
simple regression model yielded an R2 of .833 jumping to .897 with the
two predictors. The predictor intercorrelations are:

Net

Income Dividends Gain/Loss

Net - .682 .092

Income

Dividends .682 - -.522

Gain/Loss .092 -.522 -

An examination of the predictor intercorrelations reveals that Gain/Loss

and Net Income have very little correlation, but Net Income and
Dividends have a correlation of .682 and Dividends and Gain/Loss have
a correlation of -.522. These correlations might suggest
multicollinearity.

16.22 The intercorrelations of the predictor variables are:

Natural Fuel
 P a g e | 631

Gas Oil Gasoline

Natural

Gas - .570 .701

Fuel Oil .570 - .934

Gasoline .701 .934 -

Each of these intercorrelations is not small. Of particular concern is the

correlation between fuel oil and gasoline, which is .934. These two
variables seem to be adding about the same predictability to the
model. In the stepwise regression analysis only natural gas entered
the procedure. Perhaps the overlapping information between natural
gas and fuel oil and gasoline was such that fuel oil and gasoline did not
have significant unique variance to add to the prediction.
 P a g e | 632

16.23 The regression model is:

ŷ
= 564 - 27.99 x1 - 6.155 x2 - 15.90 x3

F = 11.32 with p = .003, se = 42.88, R2 = .809, adjusted R2 = .738. For

x1, t = -0.92 with p = .384, for x2, t = -4.34 with p = .002, for x3, t =
-0.71 with p = .497. Thus, only one of the three predictors, x2, is a
significant predictor in this model. This model has very good
predictability (R2 = .809). The gap between R2 and adjusted R2
underscores the fact that there are two nonsignificant predictors in this
model. x1 is a nonsignificant indicator variable.

16.24 The stepwise regression process included two steps. At step 1, x1

entered the procedure producing the model:

ŷ
= 1540 + 48.2 x1.

The R2 at this step is .9112 and the t ratio is 11.55. At step 2, x12
entered the procedure and x1 remained in the analysis. The stepwise
regression procedure stopped at this step and did not proceed. The
final model was:

ŷ
= 1237 + 136.1 x1 - 5.9 x12.
 P a g e | 633

The R2 at this step was .9723, the t ratio for x1 was 7.89, and the t ratio
for x12 was - 5.14.

16.25 In this model with x1 and the log of x1 as predictors, only the log x1 was
a significant predictor of y. The stepwise procedure only went to step
1. The regression model was:

ŷ
= - 13.20 + 11.64 Log x1. R2 = .9617 and the t ratio of Log x1
was 17.36. This model has very strong predictability using only the
log of the x1 variable.
 P a g e | 634

16.26 The regression model is:

Grain = - 4.675 + 0.4732 Oilseed + 1.18 Livestock

The value of R2 was .901 and adjusted R2 = .877.

se = 1.761. F = 36.55 with p = .000.

toilseed = 3.74 with p = .006 and tlivestock = 3.78 with p = .005. Both
predictors are significant at  = .01. This is a model with strong
predictability.

16.27 The stepwise regression procedure only used two steps. At step
2
1, Silver was the lone predictor. The value of R was .5244. At
step 2, Aluminum entered the model and Silver remained in the
model. However, the R2 jumped to .8204. The final model at step 2
was:

Gold = - 50.19 + 18.9 Silver +3.59 Aluminum.

The t values were: tSilver = 5.43 and tAluminum = 3.85.

Copper did not enter into the process at all.

 P a g e | 635

16.28 The regression model was:

Employment = 71.03 + 0.4620 NavalVessels + 0.02082 Commercial

F = 1.22 with p = .386 (not significant)

R2 = .379 and adjusted R2 = .068

The low value of adjusted R2 indicates that the model has very low
predictability. Both t values are not significant (tNavalVessels = 0.67 with

p = .541 and tCommercial = 1.07 with p = .345). Neither predictor is a

significant predictor of employment.
 P a g e | 636

16.29 There were four predictor variables. The stepwise regression

procedure went three steps. The predictor, apparel, never entered in
the stepwise process. At step 1, food entered the procedure producing
a model with an R2 of .84. At step 2, fuel oil entered and food
remained. The R2 increased to .95. At step 3, shelter entered the
procedure and both fuel oil and food remained in the model. The R2 at
this step was .96. The final model was:

All = -1.0615 + 0.474 Food + 0.269 Fuel Oil + 0.249 Shelter

The t ratios were: tfood = 8.32, tfuel oil = 2.81, tshelter = 2.56.

16.30 The stepwise regression process with these two independent variables
only went one step. At step 1, Soybeans entered in producing the
model,

Corn = - 2,962 + 5.4 Soybeans. The R2 for this model was .7868.

The t ratio for Soybeans was 5.43. Wheat did not enter in to the
analysis.

16.31 The regression model was:

 P a g e | 637

Grocery = 76.23 + 0.08592 Housing + 0.16767 Utility

+ 0.0284 Transportation - 0.0659 Healthcare

F = 2.29 with p = .095 which is not significant at  = .05.

se = 4.416, R2 = .315, and adjusted R2 = .177.

Only one of the four predictors has a significant t ratio and that is
Utility with t = 2.57 and p = .018. The ratios and their respective
probabilities are:

thousing = 1.68 with p = .109, ttransportation = 0.17 with p = .87, and

thealthcare = - 0.64 with p = .53.

This model is very weak. Only the predictor, Utility, shows much
promise in accounting for the grocery variability.
 P a g e | 638

16.32 The output suggests that the procedure only went two steps.

At step 1, x1 entered the model yielding an R2 of .7539. At step 2,

x2 entered the model and x1 remained. The procedure stopped here

with a final model of:

ŷ
= 124.5 - 43.4 x1 + 1.36 x2

The R2 for this model was .8059 indicating relatively strong

predictability with two independent variables. Since there were four
predictor variables, two of the variables did not enter the stepwise
process.

16.33 Of the three predictors, x2 is an indicator variable. An examination of

the stepwise regression output reveals that there were three steps and
that all three predictors end up in the final model. Variable x3 is the
strongest individual predictor of y and entered at step one resulting in
an R2 of .8124. At step 2, x2 entered the process and variable x3
remained in the model. The R2 at this step was .8782. At step 3,
variable x1 entered the procedure. Variables x3 and x2 remained in the
model. The final R2 was .9407. The final model was:

ŷ
= 87.89 + 0.071 x3 - 2.71 x2 - 0.256 x1
 P a g e | 639

16.34 The R2 for the full model is .321. After dropping out variable, x3, the R2
is still .321. Variable x3 added virtually no information to the model.
This is underscored by the fact that the p-value for the t test of the
slope for x3 is .878 indicating that there is no significance. The
standard error of the estimate actually drops slightly after x3 is
removed from the model.

Chapter 17
Time-Series Forecasting and Index Numbers

LEARNING OBJECTIVES

This chapter discusses the general use of forecasting in business, several

tools that are available for making business forecasts, and the nature of time
series data, thereby enabling you to:
 P a g e | 640

1. Gain a general understanding time series forecasting techniques.

2. Understand the four possible components of time-series data.

3. Understand stationary forecasting techniques.

4. Understand how to use regression models for trend analysis.

5. Learn how to decompose time-series data into their various elements

and to

forecast by using decomposition techniques

6. Understand the nature of autocorrelation and how to test for it.

7. Understand autoregression in forecasting.

CHAPTER TEACHING STRATEGY

Time series analysis attempts to determine if there is something

inherent in the history of a variable that can be captured in a way that will
help business analysts forecast the future values for the variable.

The first section of the chapter contains a general discussion about the
various possible components of time-series data. It creates the setting
against which the chapter later proceeds into trend analysis and seasonal
effects. In addition, two measurements of forecasting error are presented so
 P a g e | 641

that students can measure the error of forecasts produced by the various
techniques and begin to compare the merits of each.

A full gamut of time series forecasting techniques has been presented

beginning with the most naïve models and progressing through averaging
models and exponential smoothing. An attempt is made in the section on
exponential smoothing to show the student, through algebra, why it is called
by that name. Using the derived equations and a few selected values for
alpha, the student is shown how past values and forecasts are smoothed in
the prediction of future values. The more advanced smoothing techniques
are briefly introduced in later sections but are explained in much greater
detail on WileyPLUS.

Trend is solved for next using the time periods as the predictor
variable. In this chapter both linear and quadratic trends are explored and
compared. There is a brief introduction to Holt’s two-parameter exponential
smoothing method that includes trend. A more detailed explanation of Holt’s
method is available on WileyPLUS. The trend analysis section is placed
earlier in the chapter than seasonal effects because finding seasonal effects
makes more sense when there are no trend effects in the data or the trend
effect has been removed.

Section 17.4 includes a rather classic presentation of time series

decomposition only it is done on a smaller set of data so as not to lose the
reader. It was felt that there may be a significant number of instructors who
want to show how a time series of data can be broken down into the
components of trend, cycle, and seasonality. This text assumes a
multiplicative model rather than an additive model. The main example used
throughout this section is a database of 20 quarters of actual data on
Household Appliances. A graph of these data is presented both before and
after deseasonalization so that the student can visualize what happens when
the seasonal effects are removed. First, 4-quarter centered moving averages
are computed which dampen out the seasonal and irregular effects leaving
trend and cycle. By dividing the original data by these 4-quarter centered
moving averages (trendcycle), the researcher is left with seasonal effects
and irregular effects. By casting out the high and low values and averaging
the seasonal effects for each quarter, the irregular effects are removed.
 P a g e | 642

In regression analysis involving data over time, autocorrelation can be

a problem. Because of this, section 17.5 contains a discussion on
autocorrelation and autoregression. The Durbin-Watson test is presented as a
mechanism for testing for the presence of autocorrelation. Several possible
ways of overcoming the autocorrelation problem are presented such as the
addition of independent variables, transforming variables, and autoregressive
models.

The last section in this chapter is a classic presentation of Index

Numbers. This

section is essentially a shortened version of an entire chapter on Index

Numbers. It includes most of the traditional topics of simple index numbers,
unweighted aggregate price index numbers, weighted price index numbers,
Laspeyres price indexes, and Paasche price indexes.
 P a g e | 643

CHAPTER OUTLINE

17.1 Introduction to Forecasting

Time Series Components

The Measurement of Forecasting Error

Error

Mean Absolute Deviation (MAD)

Mean Square Error (MSE)

17.2 Smoothing Techniques

Naïve Forecasting Models

Averaging Models

Simple Averages

Moving Averages

Weighted Moving Averages

Exponential Smoothing

17.3 Trend Analysis

Linear Regression Trend Analysis

Regression Trend Analysis Using Quadratic Models

Holt’s Two-Parameter Exponential Smoothing Method

17.4 Seasonal Effects

Decomposition

Finding Seasonal Effects with the Computer

 P a g e | 644

Winters’ Three-Parameter Exponential Smoothing Method

17.5 Autocorrelation and Autoregression

Autocorrelation

Ways to Overcome the Autocorrelation Problem

Addition of Independent Variables

Transforming Variables

Autoregression

17.6 Index Numbers

Simple Index Numbers

Unweighted Aggregate Price Indexes

Weighted Price Index Numbers

Laspeyres Price Index

Paasche Price Index

 P a g e | 645

KEY TERMS

Autocorrelation Moving Average

Autoregression Naïve Forecasting Methods

Averaging Models Paasche Price Index

Cycles Seasonal Effects

Cyclical Effects Serial Correlation

Decomposition Simple Average

Deseasonalized Data Simple Average Model

Durbin-Watson Test Simple Index Number

Error of an Individual Forecast Smoothing Techniques

Exponential Smoothing Stationary

First-Difference Approach Time-Series Data

Forecasting Trend

Forecasting Error Unweighted Aggregate Price

Index Number Index Number

Irregular Fluctuations Weighted Aggregate Price

Laspeyres Price Index Index Number

Mean Absolute Deviation (MAD) Weighted Moving Average

Mean Squared Error (MSE)

SOLUTIONS TO PROBLEMS IN CHAPTER 17

 P a g e | 646

e
17.1 Period e e2

1 2.30 2.30 5.29

2 1.60 1.60 2.56

3 -1.40 1.40 1.96

4 1.10 1.10 1.21

5 0.30 0.30 0.09

6 -0.90 0.90 0.81

7 -1.90 1.90 3.61

8 -2.10 2.10 4.41

9 0.70 0.70 0.49

Total -0.30 12.30 20.43

e 
12.30
no. forecasts 9
MAD = = 1.367

e 2


20.43
no. forecasts 9
MSE = = 2.27
 P a g e | 647

e
17.2 Period Value F e e2

1 202 -

2 191 202 -11 11 121

3 173 192 -19 19 361

4 169 181 -12 12 144

5 171 174 -3 3 9
6 175 172 3 3 9
7 182 174 8 8 64
8 196 179 17 17 289

9 204 189 15 15 225

10 219 198 21 21 441

11 227 211 16 16 256

Total 35 125 1919

e 
125.00
no. forecasts 10
MAD = = 12.5

e 2


1,919
no. forecasts 10
MSE = = 191.9
 P a g e | 648

e
17.3 Period Value F e e2

1 19.4 16.6 2.8 2.8 7.84

2 23.6 19.1 4.5 4.5 20.25

3 24.0 22.0 2.0 2.0 4.00

4 26.8 24.8 2.0 2.0 4.00

5 29.2 25.9 3.3 3.3 10.89

6 35.5 28.6 6.9 6.9 47.61

Total 21.5 21.5 94.59

 e

2 1 .5
N o .F o r e c a s t s 6
MAD = = 3.583

 e2

9 4 .5 9
N o .F o r e c a s t s 6
MSE = = 15.765

e
17.4 Year Acres Forecast e e2

1 140,000 - - - -

2 141,730 140,000 1730 1730 2,992,900

3 134,590 141,038 -6448 6448 41,576,704

4 131,710 137,169 -5459 5459 29,800,681

 P a g e | 649

5 131,910 133,894 -1984 1984 3,936,256

6 134,250 132,704 1546 1546 2,390,116

7 135,220 133,632 1588 1588 2,521,744

8 131,020 134,585 -3565 3565 12,709,225

9 120,640 132,446 -11806 11806

139,381,636

10 115,190 125,362 -10172 10172

103,469,584

11 114,510 119,259 -4749 4749

22,553,001

Total -39,319 49047 361,331,847

 e

4 9 ,0 4 7
N o .F o r e c a s t s 10
MAD = = 4,904.7

 e2

3 6 1 , 3 3 1 ,8 4 7
N o .F o r e c a s t s 10
MSE = = 36,133,184.7

17.5 a.) 4-mo. mov. avg. error

44.75 14.25

52.75 13.25

61.50 9.50

64.75 21.25
 P a g e | 650

70.50 30.50

81.00 16.00

b.) 4-mo. wt. mov. avg. error

53.25 5.75

56.375 9.625

62.875 8.125

67.25 18.75

76.375 24.625

89.125 7.875
 P a g e | 651

c.) difference in errors

14.25 - 5.75 = 8.5

3.626

1.375

2.5

5.875

8.125

In each time period, the four-month moving average produces greater

errors of forecast than the four-month weighted moving average.

17.6 Period Value F( =.1) Error F( =.8) Error

Difference

1 211

2 228 211 211

3 236 213 23 225 11 12

4 241 215 26 234 7 19

5 242 218 24 240 2 22

6 227 220 7 242 -15 22

7 217 221 -4 230 -13 9

8 203 221 -18 220 -17 0

 P a g e | 652

Using alpha of .1 produced forecasting errors that were larger than

those using alpha = .8 for the first three forecasts. For the next two
forecasts (periods 6

and 7), the forecasts using alpha = .1 produced smaller errors. Each
exponential smoothing model produced nearly the same amount of
error in forecasting the value for period 8. There is no strong argument
in favor of either model.

17.7 Period Value  =.3 Error  =.7 Error 3-mo.avg.
Error

1 9.4

2 8.2 9.4 -1.2 9.4 -1.2

3 7.9 9.0 -1.1 8.6 -0.7

4 9.0 8.7 0.3 8.1 0.9 8.5

0.5

5 9.8 8.8 1.0 8.7 1.1 8.4

1.4

6 11.0 9.1 1.9 9.5 1.5 8.9

1.1

7 10.3 9.7 0.6 10.6 -0.3 9.9

0.4

8 9.5 9.9 -0.4 10.4 -0.9 10.4

-0.9

9 9.1 9.8 -0.7 9.8 -0.7 10.3

-1.2
 P a g e | 653

17.8 (a) (c) (b) (c)

Year Orders F(a) e(a) F(b) e(b)

1 2512.7

2 2739.9

3 2874.9

4 2934.1

5 2865.7

6 2978.5 2785.46 193.04 2852.36 126.14

7 3092.4 2878.62 213.78 2915.49 176.91

8 3356.8 2949.12 407.68 3000.63 356.17

9 3607.6 3045.50 562.10 3161.94 445.66

10 3749.3 3180.20 569.10 3364.41 384.89

11 3952.0 3356.92 595.08 3550.76 401.24

12 3949.0 3551.62 397.38 3740.97 208.03

13 4137.0 3722.94 414.06 3854.64 282.36

e e
17.9 Year No.Issues F(=.2) F(=.9)

1 332 -

2 694 332.0 362.0 332.0 362.0

3 518 404.4 113.6 657.8 139.8

4 222 427.1 205.1 532.0 310.0

 P a g e | 654

5 209 386.1 177.1 253.0 44.0

6 172 350.7 178.7 213.4 41.4

7 366 315.0 51.0 176.1 189.9

8 512 325.2 186.8 347.0 165.0

9 667 362.6 304.4 495.5 171.5

10 571 423.5 147.5 649.9 78.9

11 575 453.0 122.0 578.9 3.9

12 865 477.4 387.6 575.4 289.6

13 609 554.9 54.1 836.0 227.0

e e
= 2289.9 =2023.0

2289.9
12
For  = .2, MAD = = 190.8

2023.0
12
For  = .9, MAD = = 168.6

 = .9 produces a smaller mean average error.

17.10 Simple Regression Trend Model:

ŷ
= 37,969 + 9899.1 Period
 P a g e | 655

F = 1603.11 (p = .000), R2 = .988, adjusted R2 = .988,

se = 6,861, t = 40.04 (p = .000)

Quadratic Regression Trend Model:

ŷ
= 35,769 + 10,473 Period - 26.08 Period2

F = 772.71 (p = .000), R2 = .988, adjusted R2 = .987

se = 6,988, tperiod = 9.91 (p = .000), tperiodsq = -0.56 (p = .583)

The simple linear regression trend model is superior, the period 2

variable is not a significant addition to the model.

17.11 Trend line: Members = 148,141.8 – 66.0143 Year

R2 = 83.0% se = 190.1374 F = 92.82, reject the null

hypothesis.
 P a g e | 656

Year Line Fit Plot

17500

17000

16500

16000
Union Members

15500

15000

14500

1980 1985 1990 1995 2000 2005 2010

Year
 P a g e | 657

17.12

Year Line Fit Plot

1200

1000

800

600
Loans

400

200

0 1 2 3 4 5 6 7 8 9 10

Year

Trend Model:

Loans = 809.2611 + 23.18333 Year

R2 = 33.0 adjusted R2 = 23.4 se = 96.80

t = 1.86 (p = .106) F = 3.44 (p = .1.06)

Quadratic Model:

Loans = 561.1738 + 158.5037 Year – 13.5320 Year2

R2 = 90.6 adjusted R2 = 87.5 se = 39.13

tyear = 6.93 (p = .0004)

tyearsq = -6.07 (p = .000)

 P a g e | 658

F = 28.95 (p = .0008)

The graph indicates a quadratic fit rather than a linear fit. The
quadratic model produced an R2 = 90.6 compared to R2 = 33.0 for
linear trend indicating a much better fit for the quadratic model. In
addition, the standard error of the estimate drops from 96.80 to 39.13
with the quadratic model along with the t values becoming significant.
 P a g e | 659

17.13

Month Broccoli 12-Mo. Mov.Tot. 2-Yr.Tot. TC SI

Jan.(yr. 1) 132.5

Feb. 164.8

Mar. 141.2

Apr. 133.8

May 138.4

June 150.9

1655.2

July 146.6 3282.8 136.78

93.30

1627.6

Aug. 146.9 3189.7 132.90

90.47

1562.1

Sept. 138.7 3085.0 128.54

92.67

1522.9

Oct. 128.0 3034.4 126.43

98.77

1511.5

Nov. 112.4 2996.7 124.86

111.09

1485.2

Dec. 121.0 2927.9 122.00

100.83
 P a g e | 660

1442.7

Jan.(yr. 2) 104.9 2857.8 119.08

113.52

1415.1

Feb. 99.3 2802.3 116.76

117.58

1387.2

Mar. 102.0 2750.6 114.61

112.36

1363.4

Apr. 122.4 2704.8 112.70

92.08

1341.4

May 112.1 2682.1 111.75

99.69

1340.7

June 108.4 2672.7 111.36

102.73

1332.0

July 119.0

Aug. 119.0

Sept. 114.9

Oct. 106.0

Nov. 111.7

Dec. 112.3
 P a g e | 661

17.14

Month Ship 12m tot 2yr tot TC SI TCI T

C

Jan(Yr1) 1891 1952.50 2042.72

Feb 1986 1975.73 2049.87

Mar 1987 1973.78 2057.02

Apr 1987 1972.40 2064.17

May 2000 1976.87 2071.32

June 2082 1982.67 2078.46

23822

July 1878 47689 1987.04 94.51 1970.62

2085.61 94.49

23867

Aug 2074 47852 1993.83 104.02 2011.83

2092.76 96.13

23985

Sept 2086 48109 2004.54 104.06 2008.47

2099.91 95.65

24124

Oct 2045 48392 2016.33 101.42 1969.76

2107.06 93.48
 P a g e | 662

24268

Nov 1945 48699 2029.13 95.85 2024.57

2114.20 95.76

24431

Dec 1861 49126 2046.92 90.92 2002.80

2121.35 94.41

24695

Jan(Yr2) 1936 49621 2067.54 93.64 1998.97 2128.50

93.91

24926

Feb 2104 49989 2082.88 101.01 2093.12

2135.65 98.01

25063

Mar 2126 50308 2096.17 101.42 2111.85

2142.80 98.56

25245

Apr 2131 50730 2113.75 100.82 2115.35 2149.94

98.39

25485

May 2163 51132 2130.50 101.53 2137.99 2157.09

99.11

25647

June 2346 51510 2146.25 109.31 2234.07 2164.24

103.23

25863

July 2109 51973 2165.54 97.39 2213.01 2171.39

101.92

26110
 P a g e | 663

Aug 2211 52346 2181.08 101.37 2144.73

2178.54 98.45

26236

Sept 2268 52568 2190.33 103.55 2183.71

2185.68 99.91

26332

Oct 2285 52852 2202.17 103.76 2200.93 2192.83

100.37

26520

Nov 2107 53246 2218.58 94.97 2193.19

2199.98 99.69

26726

Dec 2077 53635 2234.79 92.94 2235.26

2207.13 101.27 26909

Jan(Yr3) 2183 53976 2249.00 97.07 2254.00 2214.28

101.79 27067

Feb 2230 54380 2265.83 98.42 2218.46

2221.42 99.87 27313

Mar 2222 54882 2286.75 97.17 2207.21

2228.57 99.04 27569

Apr 2319 55355 2306.46 100.54 2301.97 2235.72

102.96 27786

May 2369 55779 2324.13 101.93 2341.60 2242.87

104.40 27993

June 2529 56186 2341.08 108.03 2408.34

2250.02 107.04

28193

July 2267 56539 2355.79 96.23 2378.80 2257.17

105.39

28346
 P a g e | 664

Aug 2457 56936 2372.33 103.57 2383.35

2264.31 105.26

28590

Sept 2524 57504 2396.00 105.34 2430.19

2271.46 106.99

28914

Oct 2502 58075 2419.79 103.40 2409.94 2278.61

105.76

29161

Nov 2314 58426 2434.42 95.05 2408.66 2285.76

105.38

29265

Dec 2277 58573 2440.54 93.30 2450.50

2292.91 106.87

29308

Jan(Yr4) 2336 58685 2445.21 95.53 2411.98 2300.05

104.87

29377

Feb 2474 58815 2450.63 100.95 2461.20 2307.20

106.67

29438

Mar 2546 58806 2450.25 103.91 2529.06 2314.35

109.28

29368

Apr 2566 58793 2449.71 104.75 2547.15 2321.50

109.72

29425

May 2473 58920 2455.00 100.73 2444.40 2328.65

104.97
 P a g e | 665

29495

June 2572 59018 2459.08 104.59 2449.29 2335.79

104.86

29523

July 2336 59099 2462.46 94.86 2451.21 2342.94

104.62

29576

Aug 2518 59141 2464.21 102.18 2442.53

2350.09 103.93

29565

Sept 2454 59106 2462.75 99.64 2362.80

2357.24 100.24

29541

Oct 2559 58933 2455.54 104.21 2464.84

2364.39 104.25

29392

Nov 2384 58779 2449.13 97.34 2481.52

2371.53 104.64

29387

Dec 2305 58694 2445.58 94.25 2480.63

2378.68 104.29

29307

Jan(Yr5) 2389 58582 2440.92 97.87 2466.70 2385.83

103.39

29275

Feb 2463 58543 2439.29 100.97 2450.26

2392.98 102.39

29268
 P a g e | 666

Mar 2522 58576 2440.67 103.33 2505.22

2400.13 104.38

29308

Apr 2417 58587 2441.13 99.01 2399.25 2407.27

99.67

29279

May 2468 58555 2439.79 101.16 2439.46 2414.42

101.04

29276

June 2492 58458 2435.75 102.31 2373.11

2421.57 98.00

29182

July 2304 58352 2431.33 94.76 2417.63 2428.72

99.54

29170

Aug 2511 58258 2427.42 103.44 2435.74

2435.87 99.99

29088

Sept 2494 57922 2413.42 103.34 2401.31

2443.01 98.29

28834

Oct 2530 57658 2402.42 105.31 2436.91

2450.16 99.46

28824

Nov 2381 57547 2397.79 99.30 2478.40

2457.31 100.86

28723

Dec 2211 57400 2391.67 92.45 2379.47 2464.46

96.55
 P a g e | 667

28677

Jan(Yr6) 2377 57391 2391.29 99.40 2454.31 2471.61

99.30

28714

Feb 2381 57408 2392.00 99.54 2368.68

2478.76 95.56

28694

Mar 2268 57346 2389.42 94.92 2252.91

2485.90 90.63

28652

Apr 2407 57335 2388.96 100.76 2389.32 2493.05

95.84

28683

May 2367 57362 2390.08 99.03 2339.63 2500.20

93.58

28679

June 2446 57424 2392.67 102.23 2329.30 2507.35

92.90

28745

July 2341

Aug 2491

Sept 2452

Oct 2561

Nov 2377

Dec 2277

Seasonal Indexing:

Month Year1 Year2 Year3 Year4 Year5 Year6 Index

 P a g e | 668

Jan 93.64 97.07 95.53 97.87 99.40 96.82

Feb 101.01 98.42 100.95 100.97 99.54 100.49

Mar 101.42 97.17 103.91 103.33 94.92 100.64

Apr 100.82 100.54 104.75 99.01 100.76 100.71

May 101.53 101.93 100.73 101.16 99.03 101.14

June 109.31 108.03 104.59 102.31 102.23 104.98

July 94.51 97.39 96.23 94.86 94.76 95.28

Aug 104.02 101.37 103.57 102.18 103.44 103.06

Sept 104.60 103.55 105.34 99.64 103.34 103.83

Oct 101.42 103.76 103.40 104.21 105.31 103.79

Nov 95.85 94.97 95.05 97.24 99.30 96.05

Dec 90.92 92.94 93.30 94.25 92.45 92.90

Total 1199.69

Adjust each seasonal index by 1.0002584

Final Seasonal Indexes:

Month Index

Jan 96.85

Feb 100.52

Mar 100.67

Apr 100.74

May 101.17
 P a g e | 669

June 105.01

July 95.30

Aug 103.09

Sept 103.86

Oct 103.82

Nov 96.07

Dec 92.92

Ŷ
Regression Output for Trend Line: = 2035.58 + 7.1481 X

R2 = .682, se = 102.9

Note: Trend Line was determined after seasonal effects were removed (based on

TCI column).

17.15 Regression Analysis

The regression equation is: Food = 1.4228 + 0.4925 Housing

Predictor Coef t-ratio p

Constant 1.4228 4.57 0.0001

Shelter 0.4925 7.99 0.0000

s = 0.939 R-sq = 72.7% R-sq(adj) = 71.5%

Yˆ
Food Housing e e2

8.5 15.7 9.1555 -0.6555 0.4296

 P a g e | 670

7.8 11.5 7.0868 0.7132 0.5086

4.1 7.2 4.9690 -0.8690 0.7551

2.3 2.7 2.7526 -0.4526 0.2048

3.7 4.1 3.4421 0.2579 0.0665

2.3 4.0 3.3929 -1.0929 1.1944

3.3 3.0 2.9004 0.3996 0.1597

4.0 3.0 2.9004 1.0996 1.2092

4.1 3.8 3.2944 0.8056 0.6490

5.7 3.8 3.2944 2.4056 5.7870

5.8 4.5 3.6391 2.1609 4.6693

3.6 4.0 3.3929 0.2071 0.0429

1.4 2.9 2.8511 -1.4511 2.1057

2.1 2.7 2.7526 -0.6526 0.4259

2.3 2.5 2.6541 -0.3541 0.1254

2.8 2.6 2.7033 0.0967 0.0093

3.2 2.9 2.8511 0.3489 0.1217

2.6 2.6 2.7033 -0.1033 0.0107

2.2 2.3 2.5556 -0.3556 0.1264

2.2 2.2 2.5063 -0.3063 0.0938

2.3 3.5 3.1466 -0.8466 0.7168

3.1 4.0 3.3929 -0.2929 0.0858

1.8 2.2 2.5063 -0.7063 0.4989

2.1 2.5 2.6541 -0.5541 0.3070

3.4 2.5 2.6541 0.7459 0.5564

2.5 3.3 3.0481 -0.5481 0.3004

Total 21.1603
 P a g e | 671

 (e  e t t 1 )2
= 1.873 + 2.503 + 0.173 + 0.505 + 1.825 + 2.228 +
0.490 +

0.0864 + 2.560 + 0.060 + 3.817 + 2.750 + 0.638 +

0.089 +

0.203 + 0.064 + 0.205 + 0.064 + 0.205 + 0.064 +

0.002 +

0.292 + 0.307 + 0.171 + 0.023 + 1.690 + 1.674 =

24.561

 e 2

= 21.160

 ( e t  e t 1 ) 2

2 4 .5 6 1
 e 2
2 1 .1 6 0
D = = 1.16

Critical values of D: Using 1 independent variable, n = 26, and  = .

05,

dL = 1.30 and dU = 1.46

Since D = 1.16 is less than dL, the decision is to reject the null
hypothesis.

There is significant autocorrelation.

17.16 Regression Analysis

The regression equation is: Food = 3.1286 - 0.2003 First Diff in

Housing

Predictor Coef t-ratio p

Constant 3.1286 10.30 0.000

First Diff -.2003 -1.09 0.287

 P a g e | 672

s = 1.44854 R-sq = 4.9% R-sq(adj) = 0.8%

Food Housing First Diff in Housing

8.5 15.7 -

7.8 11.5 -4.2

4.1 7.2 -4.3

2.3 2.7 -4.5

3.7 4.1 1.4

2.3 4.0 -0.1

3.3 3.0 -1.0

4.0 3.0 0.0

4.1 3.8 0.8

5.7 3.8 0.0

5.8 4.5 0.7

3.6 4.0 -0.5

1.4 2.9 -1.1

2.1 2.7 -0.2

2.3 2.5 -0.2

2.8 2.6 0.1

3.2 2.9 0.3

2.6 2.6 -0.3

2.2 2.3 -0.3

2.2 2.2 -0.1

2.3 3.5 1.3

3.1 4.0 0.5

 P a g e | 673

1.8 2.2 -1.8

2.1 2.5 0.3

3.4 2.5 0.0

2.5 3.3 0.8

Critical values of D: Using 1 independent variable, n = 25, and  = .

05,

dL = 1.29 and dU = 1.46

Since D = 1.04 is less than dL, the decision is to reject the null
hypothesis.

There is significant autocorrelation.

17.17 The regression equation is:

Failed Bank Assets = 1,379 + 136.68 Number of Failures

ŷ
for x= 150: = 21,881 (million $)

R2 = 37.9% adjusted R2 = 34.1% se = 13,833 F = 9.78, p = .

006

The Durbin Watson statistic for this model is:

D = 2.49
 P a g e | 674

The critical table values for k = 1 and n = 18 are dL = 1.16 and dU =

1.39. Since

the observed value of D = 2.49 is above dU, the decision is to fail to

reject the null

hypothesis. There is no significant autocorrelation.

ŷ
Failed Bank Assets Number of Failures e e2

8,189 11 2,882.8 5,306.2

28,155,356

104 7 2,336.1 -2,232.1

4,982,296

1,862 34 6,026.5 -4,164.5

17,343,453

4,137 45 7,530.1 -3,393.1

11,512,859

36,394 79 12,177.3 24,216.7

586,449,390

3,034 118 17,507.9 -14,473.9

209,494,371

7,609 144 21,061.7 -13,452.7

180,974,565

7,538 201 28,852.6 -21,314.6

454,312,622

56,620 221 31,586.3 25,033.7

626,687,597

28,507 206 29,536.0 - 1,029.0

1,058,894

10,739 159 23,111.9 -12,372.9

153,089,247

43,552 108 16,141.1 27,410.9

751,357,974
 P a g e | 675

16,915 100 15,047.6 1,867.4

3,487,085

2,588 42 7,120.0 - 4,532.0

20,539,127

825 11 2,882.8 - 2,057.8

4,234,697

753 6 2,199.4 - 1,446.4

2,092,139

186 5 2,062.7 - 1,876.7

3,522,152

27 1 1,516.0 - 1,489.0
2,217,144
 P a g e | 676

17.18

Failed Bank Assets Number of Failures First Diff Failures

8,189 11 -
104 7 -4

1,862 34 27

4,137 45 11

36,394 79 34

3,034 118 39

7,609 144 26

7,538 201 57

56,620 221 20

28,507 206 -15

10,739 159 -47

43,552 108 -51

16,915 100 -8

2,588 42 -58

825 11 -31

753 6 -5

186 5 -1

27 1 -4

The regression equation is:

Failed Bank Assets = 13,019 –7.3 First Diff Failures

 P a g e | 677

R2 = 0.0% adjusted R2 = 0.0% se = 18,091.7 F

= 0.00, p = .958

The Durbin Watson statistic for this model is:

D = 1.57

The critical table values for k = 1 and n = 17 are dL = 1.13 and dU =

1.38. Since

the observed value of D = 1.57 is above dU, the decision is to fail to

reject the null

hypothesis. There is no significant autocorrelation.

17.19 Starts lag1 lag2

333.0 * *
 P a g e | 678

270.4 333.0 *

281.1 270.4 333.0

443.0 281.1 270.4

432.3 443.0 281.1

428.9 432.3 443.0

443.2 428.9 432.3

413.1 443.2 428.9

391.6 413.1 443.2

361.5 391.6 413.1

318.1 361.5 391.6

308.4 318.1 361.5

382.2 308.4 318.1

419.5 382.2 308.4

453.0 419.5 382.2

430.3 453.0 419.5

468.5 430.3 453.0

464.2 468.5 430.3

521.9 464.2 468.5

550.4 521.9 464.2

529.7 550.4 521.9

556.9 529.7 550.4

606.5 556.9 529.7

670.1 606.5 556.9

745.5 670.1 606.5

756.1 745.5 670.1

 P a g e | 679

826.8 756.1 745.5

The model with 1 lag:

Housing Starts = -8.87 + 1.06 lag 1

F = 198.67 p = .000 R2 = 89.2% adjusted R2 = 88.8% se = 48.52

The model with 2 lags:

Housing Starts = 13.66 + 1.0569 lag 2

F = 72.36 p = .000 R2 = 75.9% adjusted R2 = 74.8% Se = 70.84

The model with 1 lag is the best model with a strong R2 = 89.2%. The
model

with 2 lags is relatively strong also.

17.20 The autoregression model is: Juice = 552 + 0.645 Juicelagged2

The F value for this model is 27.0 which is significant at alpha = .001.

The value of R2 is 56.2% which denotes modest predictability. The

 P a g e | 680

adjusted R2 is 54.2%. The standard error of the estimate is 216.6. The

Durbin-Watson statistic is 1.70 indicating that there is no significant
autocorrelation in this model.

17.21 Year Price a.) Index1950 b.) Index1980

1950 22.45 100.0 32.2

1955 31.40 139.9 45.0

1960 32.33 144.0 46.4

1965 36.50 162.6 52.3

1970 44.90 200.0 64.4

1975 61.24 272.8 87.8

1980 69.75 310.7 100.0

1985 73.44 327.1 105.3

1990 80.05 356.6 114.8

1995 84.61 376.9 121.3

2000 87.28 388.8 125.1

2005 89.56 398.9 128.4

 P a g e | 681

17.22 Year Patents Index

1980 66.2 66.8

1981 71.1 71.7

1982 63.3 63.9

1983 62.0 62.6

1984 72.7 73.4

1985 77.2 77.9

1986 76.9 77.6

1987 89.4 90.2

1988 84.3 85.1

1989 102.5 103.4

1990 99.1 100.0

1991 106.7 107.7

1992 107.4 108.4

1993 109.7 110.7

1994 113.6 114.8

1995 113.8 115.3

1996 121.7 122.8

1997 124.1 125.2

1998 163.1 164.6

1999 169.1 170.6

2000 176.0 177.6

2001 184.0 185.7

 P a g e | 682

2002 184.4 186.1

2003 187.0 188.7

2004 181.3 182.9

 P a g e | 683

17.23 Year

1993 1999 2005

1.53 1.40 2.17

2.21 2.15 2.51

1.92 2.68 2.60

3.38 3.10 4.00

Totals 9.04 9.33 11.28

9 .0 4
(1 0 0 )
9 .0 4
Index1993 = = 100.0

9 .3 3
(1 0 0 )
9 .0 4
Index1999 = = 103.2

1 1 .2 8
(1 0 0 )
9 .0 4
Index2005 = = 124.8

17.24 Year
 P a g e | 684

1998 1999 2000 2001 2002 2003 2004

2005 2006

1.10 1.16 1.23 1.23 1.08 1.56 1.85 2.59

2.89

1.58 1.61 1.78 1.77 1.61 1.71 1.90 2.05

2.08

1.80 1.82 1.98 1.96 1.94 1.90 1.92 1.94

1.96

7.95 7.96 8.24 8.21 8.19 8.05 8.12 8.10

8.24

Totals 12.43 12.55 13.23 13.17 12.82 13.22 13.79

14.68 15.17

1 2 .4 3
(1 0 0 )
1 3 .2 3
Index1998 = = 94.0

1 2 .5 5
(1 0 0 )
1 3 .2 3
Index1999 = = 94.9

1 3 .2 3
(1 0 0 )
1 3 .2 3
Index2000 = = 100.0

1 3 .1 7
(1 0 0 )
1 3 .2 3
Index2001 = = 99.5
 P a g e | 685

1 2 .8 2
(1 0 0 )
1 3 .2 3
Index2002 = = 100.0

1 3 .2 2
(1 0 0 )
1 3 .2 3
Index2003 = = 101.0

1 3 .7 9
(1 0 0 )
1 3 .2 3
Index2004 = = 106.4

1 4 .6 8
(1 0 0 )
1 3 .2 3
Index2005 = = 111.0

1 5 .1 7
(1 0 0 )
1 3 .2 3
Index2006 = = 114.7

17.25 Quantity Price Price Price Price

Item 2000 2000 2004 2005 2006

1 21 0.50 0.67 0.68 0.71

2 6 1.23 1.85 1.90 1.91

3 17 0.84 0.75 0.75 0.80

4 43 0.15 0.21 0.25 0.25

 P a g e | 686

P2000Q2000 P2004Q2000 P2005Q2000 P2006Q2000

10.50 14.07 14.28 14.91

7.38 11.10 11.40 11.46

14.28 12.75 12.75 13.60

6.45 9.03 10.75 10.75

Totals 38.61 46.95 49.18 50.72

 P 2004 Q 2000 46.95

 P 2000 Q 2000 38.61
(100)
Index2000 = = = 121.6

 P 2005 Q 2000 49.18

 P 2000 Q 2000 38.61
(100)
Index2001 = = = 127.4

 P 2006 Q 2000 50.72

 P 2000 Q 2000 38.61
(100)
Index2002 = = = 131.4

17.26 Price Price Quantity Price Quantity

Item 2000 2005 2005 2006 2006

 P a g e | 687

1 22.50 27.80 13 28.11 12

2 10.90 13.10 5 13.25 8

3 1.85 2.25 41 2.35 44

P2000Q2005 P2000Q2006 P2005Q2005 P2006Q2006

292.50 270.00 361.40 337.32

54.50 87.20 65.50 106.00

75.85 81.40 92.25 103.40

Totals 422.85 438.60 519.15 546.72

 P 2005Q 2005
(1 0 0 ) 519.15
 P 2000 Q 2005 422.85
(100)
Index2005 = = = 122.8

 P 2006 Q 2006
(1 0 0 ) 546.72
 P 2000 Q 2006 438.60
(100)
Index2006 = = = 124.7

17.27 a) The linear model: Yield = 9.96 - 0.14 Month

F = 219.24 p = .000 R2 = 90.9 se = .3212

 P a g e | 688

The quadratic model: Yield = 10.4 - 0.252 Month + .

2
00445 Month

F = 176.21 p = .000 R2 = 94.4% se = .2582

In the quadratic model, both t ratios are significant,

for x: t = - 7.93, p = .000 and for x2d: t = 3.61, p = .002

The linear model is a strong model. The quadratic term adds

some

predictability but has a smaller t ratio than does the linear

term.
 P a g e | 689

b) x F e

10.08 - -

10.05 - -

9.24 -
-
9.23 - -

9.69 9.65 .04

9.55 9.55 .00

9.37 9.43 .06

8.55 9.46 .91

8.36 9.29 .93

8.59 8.96 .37

7.99 8.72 .73

8.12 8.37 .25

7.91 8.27 .36

7.73 8.15 .42

7.39 7.94 .55

7.48 7.79 .31

7.52 7.63 .11

7.48 7.53 .05

7.35 7.47 .12

7.04 7.46 .42

6.88 7.35 .47

6.88 7.19 .31

7.17 7.04 .13

7.22 6.99 .23

 P a g e | 690

e
= 6.77

6.77
20
MAD = = .3385

c)
 = .3  = .7

e e
x F F

10.08 - - - -

10.05 10.08 .03 10.08 .03

9.24 10.07 .83 10.06 .82

9.23 9.82 .59 9.49 .26

9.69 9.64 .05 9.31 .38

9.55 9.66 .11 9.58 .03

9.37 9.63 .26 9.56 .19

8.55 9.55 1.00 9.43 .88

8.36 9.25 .89 8.81 .45

8.59
8.98 .39
8.50 .09
7.99 8.86 .87 8.56 .57

8.12 8.60 .48 8.16 .04

7.91 8.46 .55 8.13 .22

7.73 8.30 .57 7.98 .25

7.39 8.13 .74 7.81 .42

 P a g e | 691

7.48 7.91 .43 7.52 .04

7.52 7.78 .26 7.49 .03

7.48 7.70 .22 7.51 .03

7.35 7.63 .28 7.49 .14

7.04 7.55 .51 7.39 .35

6.88 7.40 .52 7.15 .27

6.88 7.24 .36 6.96 .08

7.17 7.13 .04 6.90 .27

7.22 7.14 .08 7.09 .13

e e
= 10.06 = 5.97

10.06 5.97
23 23
MAD=.3 = = .4374 MAD=.7 = = .2596

 = .7 produces better forecasts based on MAD.

d). MAD for b) .3385, c) .4374 and .2596. Exponential smoothing with 
= .7

produces the lowest error (.2596 from part c).

e) 4 period 8 period

TCSI moving tots moving tots TC SI

10.08

10.05

38.60
 P a g e | 692

9.24 76.81 9.60 96.25

38.21

9.23 75.92 9.49 97.26

37.71

9.69 75.55 9.44 102.65

37.84

9.55 75.00 9.38 101.81

37.16

9.37 72.99 9.12 102.74

35.83

8.55 70.70 8.84 96.72

34.87

8.36 68.36 8.55 97.78

33.49

8.59 66.55 8.32 103.25

33.06

7.99 65.67 8.21 97.32

32.61

8.12 64.36 8.05 100.87

31.75

7.91 62.90 7.86 100.64

31.15

7.73 61.66 7.71 100.26

30.51

7.39 60.63 7.58 97.49

 P a g e | 693

30.12

7.48 59.99 7.50 99.73

29.87

7.52 59.70 7.46 100.80

29.83

7.48 59.22 7.40 101.08

29.39

7.35 58.14 7.27 101.10

28.75

7.04 56.90 7.11 99.02

28.15

6.88 56.12 7.02 98.01

27.97

6.88 56.12 7.02 98.01

28.15

7.17

7.22

1st Period 102.65 97.78 100.64 100.80 98.01

2nd Period 101.81 103.25 100.26 101.08 98.01

3rd Period 96.25 102.74 97.32 97.49 101.10

4th Period 97.26 96.72 100.87 99.73 99.02

The highs and lows of each period (underlined) are eliminated and the
others are

averaged resulting in:

 P a g e | 694

Seasonal Indexes: 1st 99.82

2nd 101.05

3rd 98.64

4th 98.67

total 398.18

Since the total is not 400, adjust each seasonal index by multiplying by
400
398.18
= 1.004571 resulting in the final seasonal indexes of:

1st 100.28

2nd 101.51

3rd 99.09

4th 99.12

17.28 Year Quantity Index Number

1992 2073 100.0

1993 2290 110.5

1994 2349 113.3

1995 2313 111.6

1996 2456 118.5

1997 2508 121.1

1998 2463 118.8

1999 2499 120.5

2000 2520 121.6

2001 2529 122.0

 P a g e | 695

2002 2483 119.8

2003 2467 119.0

2004 2397 115.6

2005 2351 113.4

2006 2308 111.3

17.29 Item 2002 2003 2004 2005 2006

1 3.21 3.37 3.80 3.73 3.65

2 0.51 0.55 0.68 0.62 0.59

3 0.83 0.90 0.91 1.02 1.06

4 1.30 1.32 1.33 1.32 1.30

5 1.67 1.72 1.90 1.99 1.98

6 0.62 0.67 0.70 0.72 0.71

Totals 8.14 8.53 9.32 9.40 9.29

 P 2002
(1 0 0 ) 
8 .1 4
(1 0 0 )
 P 2002 8 .1 4
Index2002 = = 100.0

 P 2003
(1 0 0 ) 
8 .5 3
(1 0 0 )
 P 2002 8 .1 4
Index2003 = = 104.8
 P a g e | 696

 P 2004
(1 0 0 ) 
9 .3 2
(1 0 0 )
 P 2002 8 .1 4
Index2004 = = 114.5

 P 2005
(1 0 0 ) 
9 .4 0
(1 0 0 )
 P 2002 8 .1 4
Index2005 = = 115.5

 P 2006
(1 0 0 ) 
9 .2 9
(1 0 0 )
 P 2002 8 .1 4
Index2006 = = 114.1

17.30 2003 2004 2005 2006

Item P Q P Q P Q P Q

1 2.75 12 2.98 9 3.10 9 3.21 11

2 0.85 47 0.89 52 0.95 61 0.98 66

3 1.33 20 1.32 28 1.36 25 1.40 32

Laspeyres: P2003Q2003 P2006Q2003

33.00 38.52

39.95 46.06

26.60 28.00

Totals 99.55 112.58

 P a g e | 697

 P 2006 Q 2003
(1 0 0 ) 112.58
 P 2003Q 2003 99.55
(100)
Laspeyres Index2006 = = = 113.1

Paasche2005: P2003Q2005 P2005Q2005

24.75 27.90

51.85 57.95

33.25 34.00

Totals 109.85 119.85

 P 2005Q 2005
(1 0 0 ) 119.85
 P 2003Q 2005 109.85
(100)
Paasche Index2005 = = = 109.1
P a g e | 698
 P a g e | 699

17.31 a) moving average b)  = .2

e e
Year Quantity F F

1980 6559

1981 6022 6022.00

1982 6439 6022.00

1983 6396 6340.00 56.00 6105.40 290.60

1984 6405 6285.67 119.33 6163.52
241.48

1985 6391 6413.33 22.33 6211.82 179.18

1986 6152 6397.33 245.33 6247.65 95.65

1987 7034 6316.00 718.00 6228.52 805.48

1988 7400 6525.67 874.33 6389.62 1010.38

1989 8761 6862.00 1899.00 6591.69 2169.31

1990 9842 7731.67 2110.33 7025.56 2816.45

1991 10065 8667.67 1397.33 7588.84 2476.16

1992 10298 9556.00 742.00 8084.08 2213.93

1993 10209 10068.33 140.67 8526.86

1682.14

1994 10500 10190.67 309.33 8863.29

1636.71

1995 9913 10335.67 422.67 9190.63

722.37

1996 9644 10207.33 563.33 9335.10

308.90
 P a g e | 700

1997 9952 10019.00 67.00 9396.88

555.12

1998 9333 9836.33 503.33 9507.91 174.91

1999 9409 9643.00 234.00 9472.93 63.93

2000 9143 9564.67 421.67 9460.14 317.14

2001 9512 9295.00 217.00 9396.71 115.29

2002 9430 9354.67 75.33 9419.77 10.23

2003 9513 9361.67 151.33 9421.82 91.18

2004 10085 9485.00 600.00 9440.05 644.95

e e
=11,889.67 =18,621.46

e 1 1 ,8 8 9 .6 7
numberforecasts 22
MADmoving average = = = 540.44

e 1 8 , 6 2 1 .4 6
numberforecasts 22
MAD=.2 = = = 846.43

c) The three-year moving average produced a smaller MAD (540.44) than

did

exponential smoothing with  = .2 (MAD = 846.43). Using MAD as the

criterion, the three-year moving average was a better forecasting tool
than the exponential smoothing with  = .2.
 P a g e | 701

17.32-17.34

Month Chem 12m tot 2yr tot TC SI TCI T

Jan(1) 23.701

Feb 24.189

Mar 24.200

Apr 24.971

May 24.560

June 24.992

288.00

July 22.566 575.65 23.985

94.08 23.872 23.917

287.65

Aug 24.037 575.23 23.968 100.29

24.134 23.919

287.58

Sept 25.047 576.24 24.010 104.32

24.047 23.921

288.66

Oct 24.115 577.78 24.074 100.17

24.851 23.924

289.12

Nov 23.034 578.86 24.119 95.50

24.056 23.926
 P a g e | 702

289.74

Dec 22.590 580.98 24.208 93.32

23.731 23.928

291.24

Jan(2) 23.347 584.00 24.333 95.95

24.486 23.931

292.76

Feb 24.122 586.15 24.423 98.77

24.197 23.933

293.39

Mar 25.282 587.81 24.492 103.23

23.683 23.936

294.42

Apr 25.426 589.05 24.544 103.59 24.450

23.938

294.63

May 25.185 590.05 24.585 102.44

24.938 23.940

295.42

June 26.486 592.63 24.693 107.26

24.763 23.943

297.21

July 24.088 595.28 24.803 97.12

25.482 23.945

298.07

Aug 24.672 597.79 24.908 99.05

24.771 23.947

299.72
 P a g e | 703

Sept 26.072 601.75 25.073 103.98 25.031

23.950

302.03

Oct 24.328 605.59 25.233 96.41

25.070 23.952

303.56

Nov 23.826 607.85 25.327 94.07

24.884 23.955

304.29

Dec 24.373 610.56 25.440 95.81

25.605 23.957

306.27

Jan(3) 24.207 613.27 25.553 94.73

25.388 23.959

307.00

Feb 25.772 614.89 25.620 100.59

25.852 23.962

307.89

Mar 27.591 616.92 25.705 107.34

25.846 23.964

309.03

Apr 26.958 619.39 25.808 104.46

25.924 23.966

310.36

May 25.920 622.48 25.937 99.93

25.666 23.969 312.12

June 28.460 625.24 26.052 109.24

26.608 23.971
 P a g e | 704

313.12

July 24.821 627.35 26.140 94.95

26.257 23.974 314.23

Aug 25.560 629.12 26.213 97.51

25.663 23.976

314.89

Sept 27.218 631.53 26.314 103.44

26.131 23.978

316.64

Oct 25.650 635.31 26.471 96.90

26.432 23.981

318.67

Nov 25.589 639.84 26.660 95.98

26.725 23.983

321.17

Dec 25.370 644.03 26.835 94.54

26.652 23.985

322.86

Jan(4) 25.316 647.65 26.985 93.82

26.551 23.988

324.79

Feb 26.435 652.98 27.208 97.16

26.517 23.990

328.19

Mar 29.346 659.95 27.498 106.72

27.490 23.992

331.76

Apr 28.983 666.46 27.769 104.37

27.871 23.995
 P a g e | 705

334.70

May 28.424 672.57 28.024 101.43

28.145 23.997

337.87

June 30.149 679.39 28.308 106.50

28.187 24.000

341.52

July 26.746 686.66 28.611 93.48

28.294 24.002

345.14

Aug 28.966 694.30 28.929 100.13

29.082 24.004

349.16

Sept 30.783 701.34 29.223 105.34

29.554 24.007

352.18

Oct 28.594 706.29 29.429 97.16

29.466 24.009

354.11

Nov 28.762 710.54 29.606 97.14

30.039 24.011

356.43

Dec 29.018 715.50 29.813 97.33

30.484 24.014

359.07

Jan(5) 28.931 720.74 30.031 96.34

30.342 24.016

361.67
 P a g e | 706

Feb 30.456 725.14 30.214 100.80

30.551 24.019

363.47

Mar 32.372 727.79 30.325 106.75

30.325 24.021

364.32

Apr 30.905 730.25 30.427 101.57

29.719 24.023 365.93

May 30.743 733.94 30.581 100.53

30.442 24.026

368.01

June 32.794 738.09 30.754 106.63

30.660 24.028

370.08

July 29.342

Aug 30.765

Sept 31.637

Oct 30.206

Nov 30.842

Dec 31.090

Seasonal Indexing:

Month Year1 Year2 Year3 Year4 Year5 Index

Jan 95.95 94.73 93.82 96.34 95.34

Feb 98.77 100.59 97.16 100.80 99.68

Mar 103.23 107.34 106.72 106.75 106.74

Apr 103.59 104.46 104.37 101.57 103.98

 P a g e | 707

May 102.44 99.93 101.43 100.53 100.98

June 107.26 109.24 106.50 106.63 106.96

July 94.08 97.12 94.95 93.48

94.52

Aug 100.29 99.05 97.51 100.13

99.59

Sept 104.32 103.98 103.44 105.34 104.15

Oct 100.17 96.41 96.90 97.16

97.03

Nov 95.50 94.07 95.98 97.14

95.74

Dec 93.32 95.81 94.54 97.33

95.18

Total 1199.88

Adjust each seasonal index by 1200/1199.88 = 1.0001

 P a g e | 708

Final Seasonal Indexes:

Month Index

Jan 95.35

Feb 99.69

Mar 106.75

Apr 103.99

May 100.99

June 106.96

July 94.53

Aug 99.60

Sept 104.16

Oct 97.04

Nov 95.75

Dec 95.19

Regression Output for Trend Line:

ŷ
= 22.4233 + 0.144974 x

R2 = .913

Regression Output for Quadratic Trend:

 P a g e | 709

ŷ
= 23.8158 + 0.01554 x + .000247 x2

R2 = .964

In this model, the linear term yields a t = 0.66 with p = .513 but the
squared term predictor yields a t = 8.94 with p = .000.

Regression Output when using only the squared predictor term:

ŷ
= 23.9339 + 0.00236647 x2

R2 = .964

Note: The trend model derived using only the squared predictor was used in computing T
(trend) in the decomposition process.
 P a g e | 710

17.35

2004 2005 2006

Item Price Quantity Price Quantity Price

Quantity
Margarine (lb.) 1.26 21 1.32 23 1.39
22

Shortening (lb.) 0.94 5 0.97 3 1.12 4

Milk (1/2 gal.) 1.43 70 1.56 68 1.62

65

Cola (2 liters) 1.05 12 1.02 13 1.25

11

Potato Chips (12 oz.) 2.81 27 2.86 29 2.99

28

Total 7.49 7.73

8.37

 P 2004
(1 0 0 ) 
7 .4 9
(1 0 0 )
 P 2004 7 .4 9
Index2004 = = 100.0

 P 2005
(1 0 0 ) 
7 .7 3
(1 0 0 )
 P 2004 7 .4 9
Index2005 = = 103.2

 P 2006
(1 0 0 ) 
8 .3 7
(1 0 0 )
 P 2004 7 .4 9
Index2006 = = 111.8

P2004Q2004 P2005Q2004 P2006Q2004

 P a g e | 711

26.46 27.72 29.19

4.70 4.85 5.60

100.10 109.20 113.40

12.60 12.24 15.00

75.87 77.22 80.73

Totals 219.73 231.23 243.92

 P 2005Q 2004
(1 0 0 ) 2 3 1 .2 3
 P 2004 Q 2004 2 1 9 .7 3
(1 0 0 )
IndexLaspeyres2005 = = = 105.2

 P 2006 Q 2004
(1 0 0 ) 2 4 3 .9 2
 P 2004 Q 2004 2 1 9 .7 3
(1 0 0 )
IndexLaspeyres2006 = = = 111.0

P2004Q2005 P2004Q2006 P2005Q2005 P2006Q2006

28.98 27.726 30.36 30.58

2.82 3.76 2.91 4.48

97.24 92.95 106.08 105.30

13.65 11.55 13.26 13.75

81.49 78.68 82.94 83.72

Total 224.18 214.66 235.55 237.83

 P a g e | 712

 P 2005Q 2005
(1 0 0 ) 2 3 5 .5 5
 P 2004 Q 2005 2 2 4 .1 8
(1 0 0 )
IndexPaasche2005 = = = 105.1

 P 2006 Q 2006
(1 0 0 ) 2 3 7 .8 3
 P 2004 Q 2006 2 1 4 .6 6
(1 0 0 )
IndexPaasche2006 = = = 110.8

ŷ
17.36 = 9.5382 – 0.2716 x

ŷ
(7) = 7.637

R2 = 40.2% F = 12.78, p = .002

se = 0.264862

Durbin-Watson:

n = 21 k=1  = .05

D = 0.44
 P a g e | 713

dL = 1.22 and dU = 1.42

Since D = 0.44 < dL = 1.22, the decision is to reject the null

hypothesis.

There is significant autocorrelation.

 P a g e | 714

17.37 Year x Fma Fwma SEMA SEWMA

1988 118.5

1989 123.0

1990 128.5

1991 133.6

1992 137.5 125.9 128.4 134.56 82.08

1993 141.2 130.7 133.1 111.30 65.93

1994 144.8 135.2 137.3 92.16 56.25

1995 148.5 139.3 141.1 85.10 54.17

1996 152.8 143.0 144.8 96.04 63.52

1997 156.8 146.8 148.8 99.50 64.80

1998 160.4 150.7 152.7 93.61 58.68

1999 163.9 154.6 156.6 86.03 53.14

2000 169.6 158.5 160.3 123.77 86.12

2001 176.4 162.7 164.8 188.38 135.26

2002 180.3 167.6 170.3 161.93 100.80

2003 184.8 172.6 175.4 150.06 89.30

2004 189.5 177.8 180.3 137.48 85.56

2005 195.7 182.8 184.9 167.70 115.78

SE = 1,727.60 1,111.40

SE 1 7 2 7 .6 0

N o .F o r e c a sts 14
MSEma = = 123.4
 P a g e | 715

SE 1 1 1 1 .4

N o .F o r e c a sts 14
MSEwma = = 79.39

The weighted moving average does a better job of forecasting

the data using MSE as the criterion.
 P a g e | 716

17.38 The regression model with one-month lag is:

Cotton Prices = - 61.24 + 1.1035 LAG1

F = 130.46 (p = .000), R2 = .839, adjusted R2 = .833,

se = 17.57, t = 11.42 (p = .000).

The regression model with four-month lag is:

Cotton Prices = 303.9 + 0.4316 LAG4

F = 1.24 (p = .278), R2 .053, adjusted R2 = .010,

se = 44.22, t = 1.11 (p = .278).

The model with the four-month lag does not have overall significance
and has an

adjusted R2 of 1%. This model has virtually no predictability. The

model with

the one-month lag has relatively strong predictability with adjusted R2

of 83.3%. In addition, the F value is significant at  = .001 and the
standard error of the estimate is less than 40% as large as the
standard error for the four-month lag model.

17.39
 P a g e | 717

Qtr TSCI 4qrtot 8qrtot TC SI TCI T

Year1 1 54.019

2 56.495

213.574

3 50.169 425.044 53.131 94.43 51.699 53.722

211.470

4 52.891 421.546 52.693 100.38 52.341 55.945

210.076

Year2 1 51.915 423.402 52.925 98.09 52.937 58.274

213.326

2 55.101 430.997 53.875 102.28 53.063 60.709

217.671

3 53.419 440.490 55.061 97.02 55.048 63.249

222.819

4 57.236 453.025 56.628 101.07 56.641 65.895

230.206

Year3 1 57.063 467.366 58.421 97.68 58.186 68.646

237.160

2 62.488 480.418 60.052 104.06 60.177 71.503

243.258

3 60.373 492.176 61.522 98.13 62.215 74.466

248.918

4 63.334 503.728 62.966 100.58 62.676 77.534

254.810
 P a g e | 718

Year4 1 62.723 512.503 64.063 97.91 63.957 80.708

257.693

2 68.380 518.498 64.812 105.51 65.851 83.988

260.805

3 63.256 524.332 65.542 96.51 65.185 87.373

263.527

4 66.446 526.685 65.836 100.93 65.756 90.864

263.158

Year5 1 65.445 526.305 65.788 99.48 66.733 94.461

263.147

2 68.011 526.720 65.840 103.30 65.496 98.163

263.573

3 63.245 521.415 65.177 97.04 65.174 101.971

257.842

4 66.872 511.263 63.908 104.64 66.177 105.885

253.421

Year6 1 59.714 501.685 62.711 95.22 60.889 109.904

248.264

2 63.590 491.099 61.387 103.59 61.238 114.029

3 58.088

4 61.443

Quarter Year1 Year2 Year3 Year4 Year5 Year6 Index

 P a g e | 719

1 98.09 97.68 97.91 99.48 95.22

97.89

2 102.28 104.06 105.51 103.30 103.59

103.65

3 94.43 97.02 98.13 96.51 97.04

96.86

4 100.38 101.07 100.58 100.93 104.64

100.86

Total 399.26

400
399.26
Adjust the seasonal indexes by: = 1.00185343

Adjusted Seasonal Indexes:

Quarter Index

1 98.07

2 103.84

3 97.04

4 101.05

Total 400.00
 P a g e | 720

17.40 Time Period Deseasonalized Data

Q1(yr1)
55.082
Q2 54.406

Q3 51.699

Q4 52.341

Q1(yr2) 52.937

Q2 53.063

Q3 55.048

Q4 56.641

Q1(yr3) 58.186

Q2 60.177

Q3 62.215

Q4 62.676

Q1(yr4) 63.957

Q2 65.851

Q3 65.185

Q4 65.756

Q1(yr5) 66.733

Q2 65.496

Q3 65.174

Q4 66.177

Q1(yr6) 60.889

Q2 61.238

Q3 59.860

Q4 60.805
 P a g e | 721

ŷ
17.41 Linear Model: = 53.41032 + 0.532488 x

R2 = 55.7% F = 27.65 with p = .000

se = 3.43

ŷ
Quadratic Model: = 47.68663 + 1.853339 x –0.052834 x2

R2 = 76.6% F = 34.37 with p = .000

se = 2.55

In the quadratic regression model, both the linear and squared terms
have significant t statistics at alpha .001 indicating that both are
contributing. In addition, the R2 for the quadratic model is
considerably higher than the R2 for the linear model. Also, se is smaller
for the quadratic model. All of these indicate that the quadratic model
is a stronger model.
 P a g e | 722

17.42 R2 = 55.8% F = 8.83 with p = .021

se = 50.18

This model with a lag of one year has modest predictability. The
overall F is

significant at  = .05 but not at  = .01.

17.43 The regression equation is:

Equity Funds = -359.1 + 2.0898 Money Market Funds

R2 = 88.2% se = 582.685

D = 0.84

For n = 26 and  = .01, dL = 1.07 and dU = 1.22.

 P a g e | 723

Since D = 0.84 < dL = 1.07, the null hypothesis is rejected. There is

significant

autocorrelation in this model.

17.44  = .1  = .5  = .8

e e e
Year PurPwr F F F

1 6.04

2 5.92 6.04 .12 6.04 .12 6.04 .12

3 5.57 6.03 .46 5.98 .41 5.94 .37

4 5.40 5.98 .58 5.78 .38 5.64 .24

5 5.17 5.92 .75 5.59 .42 5.45 .28

6 5.00 5.85 .85 5.38 .38 5.23 .23

7 4.91 5.77 .86 5.19 .28 5.05 .14

8 4.73 5.68 .95 5.05 .32 4.94 .21

9 4.55 5.59 1.04 4.89 .34 4.77 .22

10 4.34 5.49 1.15 4.72 .38 4.59 .25

11 4.67 5.38 .71 4.53 .14 4.39 .28

12 5.01 5.31 .30 4.60 .41 4.61 .40

13 4.86 5.28 .42 4.81 .05 4.93 .07

14 4.72 5.24 .52 4.84 .12 4.87 .15

15 4.60 5.19 .59 4.78 .18 4.75 .15

16 4.48 5.13 .65 4.69 .21 4.63 .15

17 4.86 5.07 .21 4.59 .27 4.51 .35

 P a g e | 724

18 5.15 5.05 .10 4.73 .42 4.79 .36

e e e
= 10.26 . = 4.83 = 3.97

e 10.26
N 17
MAD1 = = = .60

e 4.83
N 17
MAD2 = = = .28

e 3.97
N 17
MAD3 = = = .23

The smallest mean absolute deviation error is produced using  = .

8.

The forecast for year 19 is: F(19) = (.8)(5.15) + (.2)(4.79) =

5.08
 P a g e | 725

17.45 The model is: Bankruptcies = 75,532.436 – 0.016 Year

Since R2 = .28 and the adjusted R2 = .23, this is a weak model.

et et – et-1 (et – et-1)2 et2

- 1,338.58 1,791,796

- 8,588.28 - 7,249.7 52,558,150 73,758,553

- 7,050.61 1,537.7 2,364,521 49,711,101

1,115.01 8,165.6 66,677,023 1,243,247

12,772.28 11,657.3 135,892,643 163,131,136

14,712.75 1,940.5 3,765,540 216,465,013

- 3,029.45 -17,742.2 314,785,661 9,177,567

- 2,599.05 430.4 185,244 6,755,061

622.39 3,221.4 10,377,418 387,369

9,747.30 9,124.9 83,263,800 95,009,857

9,288.84 - 458.5 210,222 86,282,549

- 434.76 - 9,723.6 94,548,397 189,016

-10,875.36 -10,440.6 109,006,128 118,273,455

- 9,808.01 1,067.4 1,139,343 96,197.060

- 4,277.69 5,530.3 30,584,218 18,298,632

- 256.80 4,020.9 16,167,637 65,946

 (e t  et 1 ) 2 e t
2

=921,525,945 =936,737,358
 P a g e | 726

 (e  e
t t 1 )2

921,525,945
e
2
t
936,737,358
D = = 0.98

For n = 16,  = .05, dL = 1.10 and dU = 1.37

Since D = 0.98 < dL = 1.10, the decision is to reject the null

hypothesis and

conclude that there is significant autocorrelation.