Chapter 13 - Correlation and Linear Regression

Chapter 13
Correlation and Linear Regression
1.
X X Y Y X X Y Y X X Y Y
2 2
X Y
4 4 -1.6 -1.8 2.56 3.24 2.88
5 6 -0.6 0.2 0.36 0.04 -0.12
3 5 -2.6 -0.8 6.76 0.64 2.08
6 7 0.4 1.2 0.16 1.44 0.48
10 7 4.4 1.2 19.36 1.44 5.28

28 29 29.2 6.8 10.6

28 29 29.2
X 5.6 Y 5.8 sx 2.7
5 5 4
6.8 10.6
sy 1.3 r 0.75
4 (5 1)(2.7)(1.3)
The 0.75 coefficient indicates a rather strong positive correlation between X and Y. (LO13-2)

2.
X X Y Y X X Y Y X X Y Y
2 2
X Y
5 13 0.13 2.38 0.02 5.64 0.30
3 15 -1.88 4.38 3.52 19.14 -8.20
6 7 1.13 -3.63 1.27 13.14 -4.08
3 12 -1.88 1.38 3.52 1.89 -2.58
4 13 -0.88 2.38 0.77 5.64 -2.08
4 11 -0.88 0.38 0.77 0.14 -0.33
6 9 1.13 -1.63 1.27 2.64 -1.83
8 5 3.13 -5.63 9.77 31.64 -17.58

39 85 20.88 79.88 -36.38

39 85 20.88
X 4.88 Y 10.63 sx 1.73
8 8 7
79.88 36.38
sy 3.38 r 0.89
7 (8 1)(1.73)(3.38)
The 0.89 indicates a very strong negative relationship between X and Y. (LO 2)

3. a. Sales

b.

13-1
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

c.
X X Y Y X X Y Y X X Y Y
2 2
X Y
1
4 5 0 -2 0 4 0
2 8 -2 -9 4 81 18
2
5 1 1 4 1 16 4
2
6 4 2 7 4 49 14
1
3 7 -1 0 1 0 0

2 8
0 5 10 150 36
20 85 10
X 4 Y 17 sx 1.58
5 5 4
150 36
sy 6.12 r 0.93
4 (5 1)(1.58)(6.12)
d. There is a strong positive association between the variables. Sales increase with the
number of advertisements. (LO13-2)

4. a.

40
Production

30
20
10
0
0 1 2 3 4 5 6
Assemblers

b. Yes, as the number of assemblers increases, so does the production.

c.
X X Y Y X X Y Y X X Y Y
2 2
X Y
2 15 -1 -9 1 81 9
4 25 1 1 1 1 1
1 10 -2 -14 4 196 28
13-2
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

5 40 2 16 4 256 32
3 30 0 6 0 36 0

15 120 10 570 70
15 120 10
X 3 Y 24 sx 1.581
5 5 4
570 70
sy 11.937 r 0.927
4 (5 1)(1.581)(11.937) (LO13-2)

5. a. Causation could go either way. Either of these variables could be the

dependent variable.
b. This scatter diagram treats crimes as the dependent variable.

c.
X X Y Y X X Y Y X X Y Y
2 2
X Y
15 17 -3.250 5.125 10.563 26.266 -16.656
17 13 -1.250 1.125 1.563 1.266 -1.406
25 5 6.750 -6.875 45.563 47.266 -46.406
27 7 8.750 -4.875 76.563 23.766 -42.656
17 7 -1.250 -4.875 1.563 23.766 6.094
12 21 -6.250 9.125 39.063 83.266 -57.031
11 19 -7.250 7.125 52.563 50.766 -51.656
22 6 3.750 -5.875 14.063 34.516 -22.031

146 95 241.500 290.875 -231.750

146 95 241.5
X 18.25 Y 11.875 sx 5.874
8 8 7
290.875 231.75
sy 6.446 r 0.874
7 (8 1)(5.874)(6.446)
d. Strong inverse relationship stated as either: As the number of police increase,
the number of crimes decrease, or as the number of crimes increase, the
number of police decrease. (LO13-2)

6. a.

13-3
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

b.
X X Y Y X X Y Y X X Y Y
2 2
X Y
9.0 8.1 0.083 1.192 0.007 1.420 0.099
7.0 6.0 -1.917 -0.908 3.674 0.825 1.741
11.0 3.6 2.083 -3.308 4.340 10.945 -6.892
12.0 4.0 3.083 -2.908 9.507 8.458 -8.967
8.0 5.0 -0.917 -1.908 0.840 3.642 1.749
7.0 10.0 -1.917 3.092 3.674 9.558 -5.926
8.0 7.6 -0.917 0.692 0.840 0.478 -0.634
11.0 8.0 2.083 1.092 4.340 1.192 2.274
10.0 8.0 1.083 1.092 1.174 1.192 1.183
12.0 6.0 3.083 -0.908 9.507 0.825 -2.801
6.0 8.6 -2.917 1.692 8.507 2.862 -4.934
6.0 8.0 -2.917 1.092 8.507 1.192 -3.184

107 82.9 54.917 42.589 -26.292

107 82.9 54.917
X 8.917 Y 6.908 sx 2.234
12 12 11
42.589 26.292
sy 1.968 r 0.544
11 (12 1)(2.234)(1.968)
c. Moderate negative correlation between age of car and selling price. (LO13-2)

0.32 12 2
t 1.07
1 (0.32) 2
7. Reject Ho if t > 1.812 Do not reject Ho. (LO13-2)

0.46 15 2
t 1.868
1 (0.46) 2
8. Reject Ho if t < 1.771 Reject Ho. (LO13-2)

9. Ho: 0 H1: > 0 Reject Ho if t > 2.552 df = 18

0.78 20 2
t 5.288
1 (0.78) 2
Reject Ho. There is a positive correlation between gallons sold and the pump price. (LO13-2)

10. Ho: 0 H1: > 0 Reject Ho if t > 1.734 df = 18

13-4
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

0.86 20 2
t 7.150
1 (0.86) 2
Reject Ho. There is a positive correlation between assets and pretax profit. (LO13-2)

11. Ho: 0 H1: > 0 Reject Ho if t > 2.650 df = 13

0.667 15 2
t 3.23
1 (0.667) 2
Reject Ho. There is a positive correlation between passengers and cost. (LO13-2)

12. a. The scatterplot reveals a moderate to strong positive relationship between the number
of beers and blood alcohol content.
Scatterplot of BAC vs Beers

0.12

0.10

0.08
BAC

0.06

0.04

0.02

1 2 3 4 5 6 7
Beers

b.

Colum Colum Colum Colum Colum

Beers BAC n1 n2 n3 n4 n5
0.0004
6 0.1 1.722 0.0211 2.97 5 0.0364
0.0001
7 0.09 2.722 0.0111 7.41 2 0.0302
0.0001
7 0.09 2.722 0.0111 7.41 2 0.0302
0.0004
4 0.1 -0.278 0.0211 0.08 5 -0.0059
0.0004
5 0.1 0.722 0.0211 0.52 5 0.0152
0.0000
3 0.07 -1.278 -0.0089 1.63 8 0.0114
0.0004
3 0.1 -1.278 0.0211 1.63 5 -0.027
0.0016
6 0.12 1.722 0.0411 2.97 9 0.0708
0.0001
6 0.09 1.722 0.0111 2.97 2 0.0191
3 0.07 -1.278 -0.0089 1.63 0.0000 0.0114

13-5
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

8
0.0008
3 0.05 -1.278 -0.0289 1.63 3 0.0369
7 0.08 2.722 0.0011 7.41 0 0.003
0.0015
1 0.04 -3.278 -0.0389 10.74 1 0.1275
0.0000
4 0.07 -0.278 -0.0089 0.08 8 0.0025
0.0003
2 0.06 -2.278 -0.0189 5.19 6 0.043
0.0016
7 0.12 2.722 0.0411 7.41 9 0.1119
0.0008
2 0.05 -2.278 -0.0289 5.19 3 0.0658
0.0034
1 0.02 -3.278 -0.0589 10.74 7 0.193

0.0127
77 1.42 77.61 8 0.7754

77 1.42
X 4.278 Y 0.0789
18 18
77.61 0.01278
sx 2.14 s y 0.0274
17 17
0.7754
r 0.779
(18 1)( 2.14)( 0.0274)
c. Ho: 0 H1: > 0 Reject Ho if t > 2.583 df = 16
0.779 18 2
t 4.97
1 (0.779) 2
Reject Ho. There is a positive correlation between beers consumed and BAC. (LO13-
2)

13. a.
X X Y Y X X Y Y X X Y Y
2 2
X Y
4 4 -1.6 -1.8 2.56 3.24 2.88
5 6 -0.6 0.2 0.36 0.04 -0.12
3 5 -2.6 -0.8 6.76 0.64 2.08
6 7 0.4 1.2 0.16 1.44 0.48
10 7 4.4 1.2 19.36 1.44 5.28

28 29 29.2 6.8 10.6

13-6
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

28 29 29.2
X 5.6 Y 5.8 sx 2.702
5 5 4
6.8 10.6
sy 1.304 r 0.752
4 (5 1)(2.702)(1.304)

b
0.752 1.304 .363
2.702 a 5.8 (.363)(5.6) 3.767
Y = 3.767 + 0.363X
b. 6.3081, found by Y = 3.7671 + 0.3630(7) (LO13-3)

14. a.
X X Y Y X X Y Y X X Y Y
2 2
X Y
5 13 0.125 2.375 0.0156 5.6406 0.297
3 15 -1.875 4.375 3.5156 19.1406 -8.203
6 7 1.125 -3.625 1.2656 13.1406 -4.078
3 12 -1.875 1.375 3.5156 1.8906 -2.578
4 13 -0.875 2.375 0.7656 5.6406 -2.078
4 11 -0.875 0.375 0.7656 0.1406 -0.328
6 9 1.125 -1.625 1.2656 2.6406 -1.828
8 5 3.125 -5.625 9.7656 31.6406 -17.578

39 85 20.875
79.875 -36.375
39 85 20.875
X 4.875 Y 10.625 sx 1.7269
8 8 7
79.875 36.375
sy 3.3780 r 0.8908
7 (8 1)(1.7269)(3.3780)

b
0.8908 3.3780 1.7425
1.7269 a 10.625 ( 1.7425)(4.875) 19.1197
Y = 19.1197 1.7425X
b. 6.9222, found by 19.1197 1.7425(7) (LO13-3)

15. a.
X X Y Y X X Y Y X X Y Y
2 2
X Y
12 9 2.9 1.6 8.41 2.56 4.64
9 7 -0.1 -0.4 0.01 0.16 0.04
14 10 4.9 2.6 24.01 6.76 12.74
6 5 -3.1 -2.4 9.61 5.76 7.44
10 8 0.9 0.6 0.81 0.36 0.54
8 6 -1.1 -1.4 1.21 1.96 1.54
10 8 0.9 0.6 0.81 0.36 0.54
10 10 0.9 2.6 0.81 6.76 2.34
5 4.0 -4.1 -3.4 16.81 11.56 13.94
7 7.0 -2.1 -0.4 4.41 0.16 0.84

13-7
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

91 74 66.90 36.40 44.60

66.9 36.4
sx 2.72641 sy 2.01108
9 9
44.6
r 0.9038
(10 1)( 2.72641)( 2.01108)
2.01108
b (0.9038) 0.66667
2.72641 a 7.4 0.66667(9.1) 1.33333
Y 1.33333 0.66667 X

b. Y 1.33333 0.66667(6) 5.333 (LO13-3)

16. a.
X X Y Y X X Y Y X X Y Y
2 2
X Y
14 24 -19.4 -37.1 376.36 1376.41 719.74
12 14 -21.4 -47.1 457.96 2218.41 1007.94
20 28 -13.4 -33.1 179.56 1095.61 443.54
16 30 -17.4 -31.1 302.76 967.21 541.14
46 80 12.6 18.9 158.76 357.21 238.14
23 30 -10.4 -31.1 108.16 967.21 323.44
48 90 14.6 28.9 213.16 835.21 421.94
50 85 16.6 23.9 275.56 571.21 396.74
55 120 21.6 58.9 466.56 3469.21 1272.24
50 110 16.6 48.9 275.56 2391.21 811.74

334 611 2814.4 14248.9 6176.6

334 611 2814.4
X 33.4 Y 61.1 sx 17.68364
10 10 9
14248.9 6176.6
sy 39.78959 r 0.9753677
9 (10 1)(17.68364)(39.789585)
(0.9753677)( 39.78959)
b 2.19465
17.68364
a 61.1 2.19465(33.4) 12.201 Y = 12.201 + 2.1946X
b. 75.58, found by Y = -12.201 + 2.1946(40) (LO13-3)

17. a.

15
Earnings

10

0
0 25 50 75 100
13-8
Sales
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

629.6
r 0.673
b. (12 1)( 26.17)( 3.25)
3.25 64.1 501.10
b 0.673 0.0836 a (0.0836) 1.8507
c. 26.17 12 12
d. Y = 1.8507 + 0.0836(50.0) = 6.0307 ($ million) (LO13-3)

18. a.

15
Return

10

0
0 200 400 600 800
Assets

116.13
r 0.04605
b. (9 1)(192.11)(1.6409)
c. There is very little association between the amount of funds a fund has and its return.
1.6409 87.9 3504.5
b 0.04605 0.0003933 a ( 0.0003933) 9.9198
d. 192.11 9 9

e. Y = 9.9198 0.0003933(400.0) = 9.7625 The equation should be used with caution.

Assets do not account for much of the variation in the rate of return. (LO13-3)

6.44621 95 146
b 0.8744 0.9596 a ( 0.9596) 29.3877
19. a. 5.87367 8 8
b. 10.1957 found by 29.3877 0.9596(20)
c. For each policeman added, crime goes down approximately by one and without any
policemen just over 29 crimes would occur on average. (LO13-3)

1.968 82.9 107

b 0.544 0.479 a (0.479) 11.179
20. a. 2.234 12 12
b. 6.389 found by 11.179 0.479(10)
c. For each additional year the car ages the value decreases $479. (LO13-3)

21. Ho: 0 H1: < 0 df = n 2 = 8 2 = 6 Reject Ho if t < 1.943

t = 0.96/0.22= 4.364 Reject Ho and conclude the slope is less than zero. (LO13-4)

22. Ho: 0 H1: < 0 df = n 2 = 12 2 = 10 Reject Ho if t < 1.812

t = 0.479/0.234 = 2.05 Reject Ho and conclude the slope is less than zero. (LO13-4)

23. Ho: = 0 H1: 0 df = n 2 = 12 2 = 10 Reject Ho if t not between 2.228 and 2.228

t =0.08/0.03 = 2.667 Reject Ho and conclude the slope is different from zero. (LO13-4)

13-9
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

24. Ho: 0 H1: < 0 df = n 2 = 9 2 = 7 Reject Ho if t < 1.895

t = 0.00039/0.0032= 0.122 Do not reject Ho and conclude the slope may be more than zero.
(LO13-4)
68.4814
25. The standard error of estimate is 3.378, found by 82 .
The coefficient of determination is 0.76, found by (0.874)2. Seventy six percent of the
variation in crimes can be explained by the variation in police. (LO13-5)

30
26. The standard error of estimate is 1.732, found by 12 2 .
The coefficient of determination is 0.296, found by (0.544)2. So, 29 percent of the variation
in the selling price is explained by the variation in the age of the car. (LO13-5)

6.667
27. The standard error of estimate is 0.913, found by 10 2 . The coefficient of determination is
0.817, found by 0.9042. Just under eighty two percent of the variation in kilowatt hours can be
explained by the variation in the number of rooms. (LO13-5)

693
28. The standard error of estimate is 9.31, found by 10 2 . The coefficient of determination is
2
0.951, found by 0.975 . Ninety five percent of the variation in sales can be explained by the
variation in the number of contacts. (LO13-5)

29. a. r2 = 1000/1500 = 0.667

b. 0.82, found by 0.667
500
s yx
c. 6.20, found by 15 2 (LO13-5)

30. Source DF SS MS
Regression 1 7200 7200
Error 18 1800 100
Total 19 9000 (LO13-5)

1 ( 7 5. 6) 2
6.308 3.182(0.993) 6.308 1.633
31. a. 5 29.2 The interval is from 4.675
up to 7.941.

1 (7 5.6) 2
6.308 3.182(0.993) 1 6.308 3.557
b. 5 29.2
This interval is from 2.751 up to 9.865. (LO13-6)

13-10
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

1 (7.0 4.875)2
6.9222 (2.447)(1.6578) 6.9222 2.37007
8 211 (39) 2 / 8)

32. a. [4.5522,9.2923]

1 (7.0 4.875) 2
6.9222 (2.447)(1.6578) 1 6.9222 4.6982
8 211 (39)2 / 8

b. [2.2238,11.6208] (LO13-6)

1 (6 9.1) 2
5.333 2.306(0.913) 5.333 1.039
33. a. 10 66.9 .
The confidence interval is from 4.294 up to 6.372.

1 (6 9.1) 2
5.333 2.306(0.913) 1 5.333 2.348
b. 10 66.9 .
The prediction interval is from 2.985 up to 7.681. (LO13-6)

1 ( 40 33.4) 2
75.5846 2.306(9.31) 75.5846 7.2959
34. a. 10 2814.4
The confidence interval is from 68.2887 up to 82.8805.
1 ( 40 33.4) 2
75.5846 2.306(9.31) 1 75.5846 22.6757
b. 10 2814.4 .
The prediction interval is from 52.9089 up to 98.2603. (LO13-6)

35.

The correlation of X and Y is 0.2975. The scatter plot reveals the variables do not appear to be
linearly related.

13-11
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

However, the correlation of X2 and Y is 0.9975 which indicates almost a perfect linear
relationship. (LO13-2)

36. a. The relationship does not appear to be linear. It is sensible that as Score increases the
Prize Amount decreases, as low scores are better in golf.

Master's Prize versus Score

$1,600,000

$1,400,000

$1,200,000

$1,000,000
Prize

$800,000

$600,000

$400,000

$200,000

$0
280 285 290 295 300 305
Score

b. The correlation is -0.6906. Thus 47.59 percent of the variation in prize amount is
explained by score.

13-12
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

Masters Log(Prize) versus Score

6.0

5.5
Log-Prize

5.0

4.5

4.0
280 285 290 295 300 305
Score

c. The regression equation is Log-Prize = 30.249 - 0.087636 Score. The excel output
from the regression is below.

d. The coefficient of determination using Log-Prize is 87.97 percent and is much higher
than the approximately 47 percent using Prize as the dependent variable. This
indicates that the log transformation is better linear fit than the raw data.
e. If a player shoots 280, his expected prize is $514,000.
[Log(Prize)=30.249-0.087636*(280)=5.711. Antilog of 5.711 is 514,000.] (LO13-7)

37. Ho: 0 H1: > 0 Reject Ho if t > 1.714

13-13
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

0.94 25 2
t 13.213
1 (0.94) 2
Reject Ho there is a positive correlation between passengers and weight of luggage. (LO13-2)

38. Ho: 0 H1: > 0 Reject Ho if t > 2.552

0.40 20 2
t 1.852
1 (0.40) 2
Do not reject Ho. We cannot conclude that there is a positive correlation between GPA and
family income. (LO13-2)

39. Ho: 0 H1: > 0 Reject Ho if t > 2.764

0.47 12 2
t 1.684
1 (0.47) 2
Do not reject Ho. There is not a positive correlation between engine size and performance.
Using the table we know the p-value is greater than 0.05, but less than 0.10. Using software
we can find the actual p-value to be about 0.0616. The p-value indicates there is a weak
amount of evidence (less than 10%), but not at a more normal significance level of 5%.
(LO13-2)

40. a. Revenue increases slightly as the number of occupied rooms increases.

Scatterplot of I ncome vs Occupied

1550

1500
Income

1450

1400

1350

10 20 30 40 50 60
Occupied

b. Pearson correlation of Income and Occupied = 0.423

c. Ho: 0 H1: > 0 Reject Ho if t > 1.319 df = 23
0.423 25 2
t 2.239
1 (0.423) 2
Reject Ho. There is a positive correlation between revenue and occupied rooms.
d. 17.9%, found by (0.423)2, of the variation in revenue is explained by variation in
occupied rooms (LO13-2)

13-14
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

41. a. The sales volume is inversely related to their market share.

b. The correlation coefficient is -0.691, there is a moderate inverse linear relationship

between the two variables.

c. Ho: 0 H1: < 0 Reject Ho if t < -2.718 df = 11

0.691 13 2
t 3.17
1 ( 0.691) 2
Reject Ho. There is a negative correlation between cars sold and market share.
d. 47.7%, found by (-0.691)2, of the variation in market share is accounted for by
variation in cars sold. (LO13-2)

42. Ho: 0 H1: < 0 Reject Ho if t < 1.697

0.363 32 2
t 2.13
1 0.1318
Reject Ho. There is a negative correlation between square feet and rental rate. (LO13-2)

43. a. The correlation is -0.3932. If a team is scoring (more points for) it is also keeping the
ball away from the opponent (fewer points against).Thus a negative correlation is not
surprising.
b. The coefficient of determination is the square of the correlation coefficient. It is
0.1546, found by squaring (-0.3932). That means 15% of the variation of the one
variable is explained by the other.
c. Ho: 0 H1: < 0 Reject Ho if t < 1.697
0.393 32 2
t 2.34
1 ( 0.393) 2
Reject Ho. There is a negative correlation between points scored and points allowed.
d. For the National conference (NFC): Ho: 0 H1: < 0 Reject Ho if t < 1.761
0.139 16 2
t 0.53
1 ( 0.139) 2
Do not reject Ho. We cannot say there is a negative correlation between points scored
and points allowed in the NFC.

13-15
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

For the American conference (AFC): Ho: 0 H1: < 0 Reject Ho if t < 1.761
0.576 16 2
t 2.64
1 ( 0.576) 2
Reject Ho. We can say there is a negative correlation between points scored and points
allowed in the AFC. (LO13-2)

44. a. There appears to be a positive linear relationship between the two variables

The Cotton Mill

Scatter Diagram

10

8
Sales (millions$)

0
0 1 2 3 4 5 6
Size (thousands of square feet)

b. The correlation coefficient is 0.658 with a R 2 of 0.433.

The relationship between the two variables is moderate.
c. Ho: 0 H 1: > 0 Reject Ho if t > 1.714
0.658 25 2
t 4.19
1 (0.658) 2
Reject Ho. There is a significant positive correlation. (LO13-2)

45. a. There is an inverse relationship between the variables. As the months owned increases
the number of hours exercised decreases.

15

10
Hours

0
0 2 4 6 8 10 12
Months

13-16
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

b.
X X Y Y X X Y Y X X Y Y
2 2
X Y
12 4 5.5 -1.8 30.25 3.24 -9.90
2 10 -4.5 4.2 20.25 17.64 -18.90
6 8 -0.5 2.2 0.25 4.84 -1.10
9 5 2.5 -0.8 6.25 0.64 -2.00
7 5 0.5 -0.8 0.25 0.64 -0.40
2 8 -4.5 2.2 20.25 4.84 -9.90
8 3 1.5 -2.8 2.25 7.84 -4.20
4 8 -2.5 2.2 6.25 4.84 -5.50
10 2 3.5 -3.8 12.25 14.44 -13.30
5 5 -1.5 -0.8 2.25 0.64 1.20

65 58 100.50 59.60 -64.00

65 58 100.5
X 6.5 Y 5.8 sx 3.342
10 10 9
59.6 64
sy 2.573 r 0.827
9 (10 1)(3.342)(2.573)
c. Ho: 0 H1: < 0 Reject Ho if t < 2.896
0.827 10 2
t 4.16
1 ( 0.827) 2
Reject Ho. We can conclude that there is a negative association between months owned
and hours exercised. (LO13-2)

46. The complete Anova table is below. Finding this first helps with the rest of the problem.
Source SS df MS F
Regression 300 1 300 54.0
Error 100 18 5.556
Total 400 19
100
s y x 2.3570
a. 18
b. R2 = 300/400 = 0.75
c. r 0.75 0.866 The sign of r is negative because the sign of b is negative. (LO13-5)

47. a.

13-17
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

Plot of Age and Population

34

33

Median age 32

31

30
0 1 2 3 4 5 6 7 8 9
Population (in millions)

b.
Population Media
X X X X Y Y Y Y X X Y Y
2 2
(millions) n age
X Y
2.833 31.5 0.3612 0.130465 -0.54 0.2916 -0.19505
1.233 30.5 -1.2388 1.534625 -1.54 2.3716 1.907752
2.144 30.9 -0.3278 0.107453 -1.14 1.2996 0.373692
3.849 31.6 1.3772 1.89668 -0.44 0.1936 -0.60597
8.214 34.2 5.7422 32.97286 2.16 4.6656 12.40315
1.448 34.2 -1.0238 1.048166 2.16 4.6656 -2.21141
1.513 30.7 -0.9588 0.919297 -1.34 1.7956 1.284792
1.297 31.7 -1.1748 1.380155 -0.34 0.1156 0.399432
1.257 32.5 -1.2148 1.475739 0.46 0.2116 -0.55881
0.93 32.6 -1.5418 2.377147 0.56 0.3136 -0.86341

24.718 320.4 43.84259 15.924 11.93418

24.718 320.4 43.84259
X 2.4718 Y 32.04 s x 2.207
10 10 9
15.924 11.93418
sy 1.330 r 0.452
9 (10 1)( 2.207)(1.330)
c. The slope of 0.272 indicates that for each increase of 1 million in the population that
the median age increases on average by 0.272 years.
d. The median age is 32.08 years, found by 31.4 + 0.272 (2.5).
e. The p-value (0.190) for the population variable is greater than, say 0.05. A test for
significance of that coefficient would fail to be rejected. In other words, it is possible
the population coefficient is zero.
f. Ho: = 0 H1: 0 Reject Ho if t is not between 1.86 and 1.86.
0.2722 0
t 1.43
df = 8 0.1901 Do not reject Ho. There may be no
relationship between age and population. (LO13-4)

48. a.

13-18
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

Emily Smith

20000

15000

Estimated Cost

10000

5000

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

b.
Estimated Age
Y Y Y Y X X X X X X Y Y
2 2

Cost (Y) (X)

$5,555 8 -4805.93 23096949 2.785714 7.760204 -13387.9
$17,888 3 7527.071 56656804 -2.21429 4.903061 -16667.1
$9,963 6 -397.929 158347 0.785714 0.617347 -312.658
$6,793 5 -3567.93 12730114 -0.21429 0.045918 764.5561
$10,774 5 413.0714 170628 -0.21429 0.045918 -88.5153
$16,310 2 5949.071 35391451 -3.21429 10.33163 -19122
$2,475 8 -7885.93 62187869 2.785714 7.760204 -21967.9
$2,808 10 -7552.93 57046730 4.785714 22.90306 -36146.2
$7,073 9 -3287.93 10810474 3.785714 14.33163 -12447.2
$8,978 8 -1382.93 1912491 2.785714 7.760204 -3852.44
$11,213 2 852.0714 726026 -3.21429 10.33163 -2738.8
$9,463 3 -897.929 806276 -2.21429 4.903061 1988.27
$15,055 2 4694.071 22034307 -3.21429 10.33163 -15088.1
10699981
$20,705 2 10344.07 4 -3.21429 10.33163 -33248.8

39072828
$145,053 73 1 112.3571 -172315
145,053 73 390,728,281
X 10,361 Y 5.214 sx 5482
14 14 13
112.3571 172,315
sy 2.94 r 0.822
13 (14 1)(5482)( 2.94)
c. The slope of 1534 indicates that for each increase of 1 year in the age of the car that
the estimated cost decreases on average by $1534.
d. $10,688, found by 18358 1534(5)

13-19
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

e. The p-value (0.000) for the age variable is less than most common significance levels,
such as 0.05 indicating a test for significance of that coefficient would be rejected. In
other words, the coefficient is different from zero.
f. Ho: = 0 H1: 0 Reject Ho if t is not between 1.782 and 1.782.
1533.6 0
t 5.01
df = 12 306.3 Reject Ho.
There is a relationship between age and cost. (LO13-4)

49. a. The regression equation is Y = 11.236 0.4667 X. This indicates there is a negative
relationship between the number of bids (X) and the winning bid (Y) and that for each
additional bidder the winning bid decreases by 0.4667 million.
The equation can be found using software or with the calculations below.
n = 15 X = 107 Y = 118.6

( x x) 2
73.73333 ( y y) 2
32.1893

73.73333 32.1893
sx 2.294922 sy 1.516323
14 14
34.41333
r 0.706381
(15 1)( 2.294922)(1.516323)
1.516323
b 0.706381 0.466727
2.294922

118.6 107
a ( 0.466727) 11.235986
15 15
b. Y = 11.235986 0.466727(7.0) = 7.968897
1 (7 7.1333) 2
7.9689 (2.160)(1.114) 1 7.9689 2.4854
15 (107) 2
837
15
c. [5.4835,10.4543]
d. R2 = 0.499. The number of bidders explains nearly 50 percent of the variation in the
amount of the bid. (LO13-6)

50. a. Y = 10.6678 +0.003X

163.6 1193.8
a (0.0030) 10.6678
15 15
b. Ho: 0 H1: > 0 df = n 2 = 15 2 = 13
At the 0.05 significance level, reject Ho if t > 1.771 t =0.00303/0.00159=1.90 Reject
Ho and conclude the slope is positive.
c. R2 = (0.466)2 = 0.2172 While he may not be pleased with the limited explanatory
power, the size of the offering explains over 20% of the uncertainty. (LO13-6)

51. a. There appears to be a relationship between the two variables. As the distance increases,
so does the shipping time.

13-20
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

b.

X X Y Y X X Y Y X X Y Y
2 2

X Y
1.4 100 0.25 3.3333 0.0625 11.1111 0.8333
1.3 110 0.15 13.3333 0.0225 177.7778 2.0000
1.2 105 0.05 8.3333 0.0025 69.4444 0.4167
1.1 120 -0.05 23.3333 0.0025 544.4444 -1.1667
656 5 -69.85 -3.4 4879.023 11.56 237.49
853 14 127.15 5.6 16167.123 31.36 712.04
646 6 -79.85 -2.4 6376.023 5.76 191.64
783 11 57.15 2.6 3266.123 6.76 148.59
610 8 -115.85 -0.4 13421.223 0.16 46.34
841 10 115.15 1.6 13259.523 2.56 184.24
785 9 59.15 0.6 3498.723 0.36 35.49
639 9 -86.85 0.6 7542.923 0.36 -52.11
762 10 36.15 1.6 1306.823 2.56 57.84
762 9 36.15 0.6 1306.823 0.36 21.69
862 7 136.15 -1.4 18536.823 1.96 -190.61
679 5 -46.85 -3.4 2194.923 11.56 159.29
835 13 109.15 4.6 11913.723 21.16 502.09
607 3 -118.85 -5.4 14125.323 29.16 641.79
665 8 -60.85 -0.4 3702.723 0.16 24.34
647 7 -78.85 -1.4 6217.323 1.96 110.39

14517 168 145306.55 138.80 3108.20

14517 168 145306.55
X 725.85 Y 8.4 sx 87.451
20 20 19
138.8 3108.2
sy 2.703 r 0.692
19 (20 1)(87.451)(2.703)
Ho: 0 H1: > 0 Reject Ho if t > 1.734
0.692 20 2
t 4.067
1 (0.692) 2
Ho is rejected. There is a positive association between
shipping distance and shipping time.

13-21
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

c. R2 = (0.692)2 = 0.479, nearly half of the variation in shipping time is explained

by shipping distance.
72.3147
d. The standard error of estimate is 2.004 = 18 .
Ro Day
w
Distance
s Y Residual Residual2
1 656 5 6.906 -1.906 3.6328
2 853 14 11.120 2.880 8.2944
3 646 6 6.692 -0.692 0.4789
4 783 11 9.622 1.378 1.8989
5 610 8 5.922 2.078 4.3181
6 841 10 10.863 -0.863 0.7448
7 785 9 9.665 -0.665 0.4422
8 639 9 6.542 2.458 6.0418
9 762 10 9.173 0.827 0.6839
10 762 9 9.173 -0.173 0.0299
11 862 7 11.312 -4.312 18.5933
12 679 5 7.398 -2.398 5.7504
13 835 13 10.735 2.265 5.1302
14 607 3 5.858 -2.858 8.1682
15 665 8 7.098 0.902 0.8136
16 647 7 6.713 0.287 0.0824
17 685 10 7.526 2.474 6.1207
18 720 8 8.275 -0.275 0.0756
19 652 6 6.820 -0.820 0.6724
20 828 10 10.585 -0.585 0.3422
72.3147
e. The regression equation is suggested.
It describes nearly half of the variation in shipping time with distance. (LO13-6)

52. a.

1300
1200
1100
1000
900
Spent

800
700
600
500
400
4000 5000 6000 7000 8000 9000 10000
Income

40(245,795,835) (273,387)(33,625)
b 0.13388
b. 40(1,987,875,615) (273,387) 2

13-22
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

33,625 273,387
a 0.13388 74.4
40 40
The regression equation is Spent = 74.4 + 0.134 Income. For each additional dollar
of income, 13.4 cents more is spent on groceries.
40 245,795,835 273,387 33,625
r 0.945
[40 1,987,875,615 273,387 ][40 30,662,885 33, 625 ]
2 2

c.
Ho: 0 H1: > 0 At the 5% level, reject Ho when t > 1.686.
0.945 40 2
t 17.8
1 0.945
2

Thus Ho is rejected. The population correlation is positive. (LO13-5)

53. a. The regression equation is: Price = 26.8 + 2.41 dividend. For each additional dollar
paid out in a dividend the per share price increases by $2.41 on average.
b. Ho: = 0 H1: 0 At the 5% level, reject Ho if t is not between 2.048
and 2.048. t = 2.408/0.328 = 7.34
Reject Ho and conclude slope is not zero.
5057.6
R2 0.658
c. 7682.7 65.8% of the variation in price is explained by the
dividend.
d. r .658 0.811 Ho: 0 H1: > 0
0.811 30 2
t 7.34
1 0.811
2

At the 5% level, reject Ho when t > 1.701.

Thus Ho is rejected. The population correlation is positive. (LO13-5)

54. a. 20, found by one more than the total degrees of freedom (19 + 1)

b. 25.884, found by the square root of mean square error

670
.

10,354
c. 0.4621, found by SSR/SStotal 22, 408 .
d. 0.6798, found by 0.4621
e. Yes, because the tvalue (3.93) is greater than the critical value (1.734). (LO13-5)

55. a. 35, found by one more than the total degrees of freedom (34 + 1)

b. 5457, found by the square root of mean square error

29,778,406
.

13,548,662,082
c. 0.9324, found by SSR/SStotal 14,531,349, 474 .
d. 0.9656, found by 0.9324
e. Yes, because the tvalue (21.33) is greater than the critical value (1.692). (LO13-5)

56. a.

13-23
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

X X Y Y X X Y Y X X Y Y
2 2
X Y
23.1 8.0 5.613 -5.825 31.500 33.931 -32.693
13.2 15.6 -4.288 1.775 18.383 3.151 -7.610
24.2 31.2 6.713 17.375 45.058 301.891 116.630
11.1 2.5 -6.388 -11.325 40.800 128.256 72.338
10.1 35.4 -7.388 21.575 54.575 465.481 -159.385
10.8 6.0 -6.688 -7.825 44.723 61.231 52.330
27.3 8.7 9.813 -5.125 96.285 26.266 -50.289
20.1 3.2 2.613 -10.625 6.825 112.891 -27.758

139.9 110.6 338.149

1133.095 -36.438
139.9 110.6 338.149
X 17.4875 Y 13.825 sx 6.950
8 8 7
1133.095 36.438
sy 12.723 r 0.059
7 (8 1)(6.950)(12.723)
The correlation of Return and Growth is -0.059.
Ho: 0 H1: > 0 Reject Ho if t > 1.943
0.059 8 2
t 0.145
1 (0.059)2
Fail to reject the null. It is not reasonable to say the
population correlation is positive.
b. The regression equation is Return = 17.9321 0.0322 Growth. Ho: 0 H1: < 0,
with Reject Ho if t < -1.943. The test statistic is -0.1444. We fail to reject the null
hypothesis. It is possible the relationship is positive.
c. Rockwell Collins has the largest residual (9.64769). (LO13-4)

57. a. The regression equation is Price = 773.09 + 1407.93 Speed.

b. The second laptop (1.6, 922) has a residual of 557.60 and is priced $557.60 below
the predicted price.
c. The correlation of Speed and Price is 0.835.
H0: 0 H1: > 0 Reject H0 if t > 1.8125
0.835 12 2
t 4.799
1 (0.835) 2
Reject H0. It is reasonable to say the population correlation is positive. (LO13-3)

58. a. The correlation of Wattage and Area is 0.9393. The link is direct.
b. H0: 0 H1: > 0 Reject H0 if t > 1.7341
0.939 20 2
t 11.584
1 (0.939) 2
Reject H0. It is quite reasonable to say the population correlation is positive!
c. The regression equation is Area = 22.5813 + 0.1489 Wattage.
d. The sixth heater is the best buy. It heats an area that is 53.47 square feet larger than
estimated by the regression equation and represents the largest residual. (LO13-3)

59. a. The correlation of Weight and Consumption is 0.987.

13-24
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

0.987 18 2
t 24.564
1 (0.987) 2
H0: 0 H1: > 0 Reject H0 if t > 1.746
Reject H0. It is quite reasonable to say the population correlation is positive!
b. The regression equation is Weight = 29.543 + 22.914 Consumption. Each additional
cup increases the estimated weight by 22.9 pounds.
c. The fourth dog has the largest residual and is a big over eater. It weighs 21 pounds
less than the regression equation would estimate. (LO13-3)

60. a. Y = 12.3601 + 4.7956 X The relationship between distance and damage is direct.
b. $36,338.10 found by 12.3601 + 4.7956(5)
c. 0.581, found by 1865/3209, 58.1% of the variation in damage is explained by
variation in distance.
d. 0.762 which is 0.581 It is positive because the slope is positive. There is a strong
direct link between the variables.
e. Ho: = 0 H1: 0 Reject Ho if t < -2.763 or t > 2.763
0.762 30 2
t 6.23
1 (0.762) 2 Reject Ho.
There is a connection between distance and fire damage. (LO13-5)

61. a. The relationship is direct. Fares increase for longer flights.

Scatter Diagram of fares and distances

350

300

250
Fare

200

150

100
0 500 1000 1500 2000 2500
Distance

b. The correlation of Distance and Fare is 0.656. Ho: 0 H1: > 0

0.656 30 2
t 4.599
1 (0.656) 2
Reject Ho if t >1.701 df = 28
Reject Ho. There is a significant positive correlation between fares and distances.
c. 43 percent, found by (0.656)2 , of the variation in fares is explained by the variation in
distance.

13-25
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

d. The regression equation is Fare = 147.08 + 0.05265 Distance. Each additional mile
adds $0.05265 to the fare. A 1500 mile flight would cost $226.06, found by $147.08 +
0.05265(1500).
e. A flight of 4218 miles is outside the range of the sampled data. So the regression
equation may not be useful. (LO13-5)

62. a. Using a statistical software package, the regression equation is Y = 64.7931 +

0.0703289 X, where size is the independent variable and price is the dependent
variable. The estimated selling price for a 2200 square foot home is 219.4293 ($000)
that is $219,429.30. The limits for a 95% confidence interval are 210.88 and 227.96
($000). For a 95% prediction interval the limits are 131.84 and 307.02 ($000). The
coefficient of determination is 13.767%, so this independent variable does not explain
much of the variation in selling price.
b. The regression equation is Y = 270.167 3.3534X. A home 20 miles from the center
of the city is estimated to sell for 203.0871 ($000). The 95% confidence interval is
from 190.267 up to 215.907 and the 95% prediction interval is 114.118 up to 292.056.
This interval is in ($000). Again, the coefficient of determination is quite low, 12.04%.
c. For selling price and distance r = 0.347. n is 105, so there are 103 degrees of
freedom. The null and the alternate hypotheses are:
Ho: 0 H1: < 0 Reject Ho if t < 1.66
0.371 105 2
t 3.755
1 ( 0.347) 2
Ho is rejected. There is a negative correlation between selling price and distance.
For selling price and area of the home, the correlation is 0.371. The hypotheses are:
Ho: 0 H1: > 0 Reject Ho if t > 1.66
0.380 105 2
t 4.055
1 (0.380) 2 Reject Ho. There is a positive correlation between
selling price and area of the home. (LO13-6)

63. a. There does seem to be a direct relationship between the variables.

13-26
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

b. Expected Attendance with a salary of 80 million is 2.2529 million, found by 1.17855

+ 0.01343 (80)
c. Increasing the salary by 30 million will increase attendance by 0.4029 million on
average, found by 0.01343 (30). This is also the difference between the expected
attendance with a salary of 110 and the expected attendance of 80 million.
d. The regression output from Excel is below.

Ho: 0 H1: > 0 df = n 2 = 30 2 = 28 Reject Ho if t > 1.701

t = 0.013429923 / 0.002100941 = 6.39233, Reject Ho and conclude the slope is
positive.
e. 0.593389 or 59.3389% of the variation in attendance is explained by variation in
salary.
f. The correlation between attendance and batting average is 0.6538.
Ho: 0 H1: > 0 At the 5% level, reject Ho if t > 1.701
0.6538 30 2
t 4.573
1 (0.6538) 2
Reject Ho.
The batting average and attendance is positively correlated.
The correlation between attendance and ERA is 0.0548. The correlation between
attendance and batting average is stronger than the correlation between attendance and
ERA.
Ho: 0 H1: < 0 At the 5% level, reject Ho if t < 1.701
0.0548 30 2
t 0.291
1 (0.0548) 2
Fail to reject Ho.

13-27
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.
Chapter 13 - Correlation and Linear Regression

The ERA and attendance could have no correlation. (LO13-5)

64. a. There seems to be a weak direct relationship between the two variables.
Maintenance = 386.8 + 9.070 Age
600

550

500
Maintenance

450

400

350

300
0 2 4 6 8 10 12 14
Age

b. The regression equation is Maintenance = 386.8 + 9.07(Age). An addition year adds

$9.07 to the maintenance cost. The estimated cost of a 10-year old bus is$477.50,
found by 386.8 + 9.07(10)
c. Ho: 0 H1: > 0 df = n 2 = 80 2 = 78 Reject Ho if t > 1.665
t = 9.07/1.956 = 4.64 Reject Ho and conclude the slope is positive. (LO13-4)

13-28
Copyright 2015 McGraw-Hill Education. All rights reserved. No reproduction or distribution without the prior written consent of
McGraw-Hill Education.