Chapter Eleven

Section 11.1

11.1 The chi-square distribution has only one parameter, called the degrees of freedom (df). The
shape of a chi-square distribution curve is skewed to the right for small df and becomes symmetric
for large df. The entire chi-square distribution curve lies to the right of the vertical axis. The chi-
square distribution assumes nonnegative values only, and these are denoted by the symbol χ2.

11.3 For df = 28 and .05 area in the right tail, χ2 = 41.337.

11.5 For an area of .990 in the left tail, the area in the right tail is 1 – .990 = .01. Hence, the value of
chi–square for df = 23 and .990 area in the left tail is the same as for df = 23 and .01 area in the
right tail. Thus, the chi-square value is χ2 = 41.638.

11.7 a. For an area of .025 in the left tail, the area in the right tail is 1 – .025 = .975. Hence, the value
of chi–square for df = 13 and .025 area in the left tail is the same as for df = 13 and .975 area in
the right tail. Thus, the chi-square value is χ2 = 5.009
b. For df = 13 and .995 area in the right tail, χ2 = 3.565.

Section 11.2

11.9 Α goodness–of–fit test compares the observed frequencies from a multinomial experiment with
expected frequencies derived from a certain pattern or theoretical distribution. The test evaluates
how well the observed frequencies fit the expected frequencies.

11.11 The expected frequency of a category is given by Ε = np where n is the sample size and p is the
probability that an element belongs to that category if the null hypothesis is true. The degrees of
freedom for a goodness–of–fit test are k – 1, where k denotes the number of possible outcomes (or
categories) for the experiment.

223
224 Chapter Eleven

11.13 Step 1: H0: The die is fair, Η1: The die is not fair.
Step 2: Since this is a multinomial experiment, use the chi–square distribution.
Step 3: k = 6, df = k – 1 = 6 – 1 = 5
For α = .05, the critical value of χ2 is 11.070.
Step 4: Note that the die will be fair if the probability of each of the six outcomes is the same,
which is 1/6.
Outcome O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε)2/Ε
1 7 1/6 10 –3 9 .900
2 12 1/6 10 2 4 .400
3 8 1/6 10 –2 4 .400
4 15 1/6 10 5 25 2.500
5 11 1/6 10 1 1 .100
6 7 1/6 10 –3 9 .900
n = 60 Sum = 5.200

χ2 = Σ (O – E)2/E = 5.200
Step 5: Do not reject H0 since 5.200 < 11.070.
Conclude that the die is fair.

11.15 Step 1: H0: The current distribution of responses is the same as that of May 2009.
Η1: The current distribution of responses differs from that of May 2009.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 4, df = k – 1 = 4 – 1 = 3
For α = .05, the critical value of χ2 is 7.815.
Step 4:
Response O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Champagne/caviar 104 .1546 92.76 11.24 126.338 1.362
Wine/chicken breasts 294 .5876 352.56 –58.56 3429.274 9.727
Beer/pigs in blanket 184 .2062 123.72 60.28 3633.678 29.370
Not sure 18 .0515 30.90 –12.90 166.410 5.385
n = 600 Sum =
45.844

χ2 = Σ (O – E)2/E = 45.844
Step 5: Reject H0 since 45.844 > 7.815.
Conclude that the current distribution of responses differs from that of May 2009.

11.17 Step 1: H0: The number of boxes sold of each of the five colors is the same.
Η1: The number of boxes sold of each of the five colors differs.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 225

Step 3: k = 5, df = k – 1 = 5 – 1 = 4
For α = .01, the critical value of χ2 is 13.277.
Step 4:
Color O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Blue 310 .20 278.8 31.2 973.44 3.492
Green 292 .20 278.8 13.2 174.24 .625
Orange 280 .20 278.8 1.2 1.44 .005
Red 216 .20 278.8 –62.8 3943.84 14.146
White 296 .20 278.8 17.2 295.84 1.061
n = 1394 Sum = 19.329

χ2 = Σ (O – E)2/E = 19.329
Step 5: Reject H0 since 19.329 > 13.277.
Conclude that the number of boxes sold of each of the five colors differs.

11.19 Step 1: H0: The orders are evenly distributed over all days of the week.
Η1: The orders are not evenly distributed over all days of the week.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 5, df = k – 1 = 5 – 1 = 4
For α = .05, the critical value of χ2 is 9.488.
Step 4:
Day O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Monday 92 .20 80 12 144 1.800
Tuesday 71 .20 80 –9 81 1.013
Wednesday 65 .20 80 – 15 225 2.813
Thursday 83 .20 80 3 9 .113
Friday 89 .20 80 9 81 1.013
n = 400 Sum = 6.752

χ2 = Σ (O – E)2/E = 6.752
Step 5: Do not reject H0 since 6.752 < 9.488.
Conclude that the orders are evenly distributed over all days of the week.

11.21 Step 1: H0: The percentage distribution of users' opinions is unchanged since the product was
redesigned.
Η1: The percentage distribution of users' opinions has changed since the product was
redesigned.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 4, df = k – 1 = 4 – 1 = 3
226 Chapter Eleven

For α = .025, the critical value of χ2 is 9.348.

Step 4:
Opinion O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Excellent 495 .53 424 71 5041 11.889
Satisfactory 255 .31 248 7 49 .198
Unsatisfactory 35 .07 56 –21 441 7.875
No opinion 15 .09 72 –57 3249 45.125
n = 800 Sum = 65.087

χ2 = Σ (O – E)2/E = 65.087
Step 5: Reject H0 since 65.087 > 9.348.
Conclude that the percentage distribution of users' opinions has changed since the product was
redesigned.

Sections 11.3 - 11.4

11.23 In a test of independence, we test the null hypothesis that two characteristics of the elements in a
given population are not related against the alternative hypothesis that the two characteristics are
related. See Example 11−6 in the text. In a test of homogeneity, we test if two (or more)
populations are similar with respect to the distribution of certain characteristics. See Example
11−8 in the text.

11.25 The minimum expected frequency for each cell should be 5. If this condition is not satisfied, we
may increase the sample size or combine some categories.

11.27 a. H0: The proportion in each row is the same for all four populations.
Η1: The proportion in each row is not the same for all four populations.
b. The expected frequencies are given in parentheses below the observed frequencies in the table
below.
Column 1 Column 2 Column 3 Column 4 Total
Row 1 24 81 60 121 286
(31.62) (63.95) (89.95) (100.49)
Row 2 46 64 91 72 273
(30.18) (61.04) (85.86) (95.92)
Row 3 20 37 105 93 255
(28.19) (57.01) (80.20) (89.59)
Total 90 182 256 286 814

c. df = (R – 1)(C – 1) = (3 – 1)(4 – 1) = 6
For α = .025, the critical value of χ2 is 14.449.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 227

(O − E ) 2 (24 − 31.62) 2 (81 − 63.95) 2 (60 − 89.95) 2 (121 − 100.49) 2

d. χ 2 = ∑ = + + +
E 31.62 63.95 89.95 100.49

(46 − 30.18) 2 (64 − 61.04) 2 (91 − 85.86) 2 (72 − 95.92) 2 (20 − 28.19) 2
+ + + + +
30.18 61.04 85.86 95.92 28.19

(37 − 57.01) 2 (105 − 80.20) 2 (93 − 89.59) 2

+ + +
57.01 80.20 89.59
= 1.836 + 4.546 + 9.972 + 4.186 + 8.293 + .144 + .308 + 5.965 + 2.379 + 7.023 + 7.669 + .130
= 52.451
e. Since 52.451 > 14.449, reject H0.

11.29 Step 1: H0: Gender and wearing or not wearing of seat belt are not related.
Η1: Gender and wearing or not wearing of seat belt are related.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(2 – 1) = 1
For α = .025, the critical value of χ2 is 5.024.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in the
table below.
Wearing seat belt Not wearing seat belt Total
Men 40 15 55
(42.9) (12.1)
Women 38 7 45
(35.1) (9.9)
Total 78 22 100

(O − E ) 2 (40 − 42.9) 2 (15 − 12.1) 2 (38 − 35.1) 2 (7 − 9.9) 2

χ2 =∑ = + + +
E 42.9 12.1 35.1 9.9
= .196 + .695 + .240 + .849 = 1.980
Step 5: Do not reject H0 since 1.980 < 5.024.
Conclude that gender and wearing or not wearing of seat belt are not related.
228 Chapter Eleven

11.31 a. Step 1: H0: Approval ratings by own party voters and president are not related.
Η1: Approval ratings by own party voters and president are related.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(2 – 1) = 1
For α = .01, the critical value of χ2 is 6.635.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Own Party Obama Bush Total
Approve 1091 1046 2137
(1068.5) (1068.5)
Not sure/ 109 154 263
disapprove (131.5) (131.5)
Total 1200 1200 2400

(O − E ) 2 (1091 − 1068.5) 2 (1046 − 1068.5) 2 (109 − 131.5) 2 (154 − 131.5) 2

χ2 =∑ = + + +
E 1068.5 1068.5 131.5 131.5
= .474 + .474 + 3.850 + 3.850 = 8.648
Step 5: Reject H0 since 8.648 > 6.635.
Conclude that approval ratings by own party voters and president are related.
b. Step 1: H0: Approval ratings by opposition party voters and president are not related.
Η1: Approval ratings by opposition party voters and president are related.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(2 – 1) = 1
For α = .01, the critical value of χ2 is 6.635.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Opposition Party Obama Bush Total
Approve 335 430 765
(382.5) (382.5)
Not sure/ 865 770 1635
disapprove (817.5) (817.5)
Total 1200 1200 2400

(O − E ) 2 (335 − 382.5) 2 (430 − 382.5) 2 (865 − 817.5) 2 (770 − 817.5) 2

χ2 =∑ = + + +
E 382.5 382.5 817.5 817.5
= 5.899 + 5.899 + 2.760 + 2.760 = 17.318
Step 5: Reject H0 since 17.318 > 6.635.
Conclude that approval ratings by opposition party voters and president are related.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 229

11.33 Step 1: H0: Region and causes of fire are unrelated.

Η1: Region and causes of fire are related.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(4 – 1) = 3
For α = .05, the critical value of χ2 is 7.815.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Region Arson Accident Lightning Unknown Total
A 6 9 6 10 31
(5.30) (9.38) (8.57) (7.75)
B 7 14 15 9 45
(7.70) (13.62) (12.43) (11.25)
Total 13 23 21 19 76

(O − E ) 2 (6 − 5.30) 2 (9 − 9.38) 2 (6 − 8.57) 2 (10 − 7.75) 2 (7 − 7.70) 2

χ2 =∑ = + + + +
E 5.30 9.38 8.57 7.75 7.70

(14 − 13.62) 2 (15 − 12.43) 2 (9 − 11.25) 2

+ + +
13.62 12.43 11.25
= .092 + .015 + .771 + .653 + .064 + .011 + .531 + .450 = 2.587
Step 5: Do not reject H0 since 2.587 < 7.815.
Conclude that region and causes of fire are unrelated.

11.35 Step 1: H0: The two drugs are similar in curing the patients.
Η1: The two drugs are not similar in curing the patients.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(2 – 1) = 1
For α = .01, the critical value of χ2 is 6.635.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Cured Not Cured Total
Drug I 44 16 60
(37.2) (22.8)
Drug II 18 22 40
(24.8) (15.2)
Total 62 38 100

(O − E ) 2 (44 − 37.2) 2 (16 − 22.8) 2 (18 − 24.8) 2 (22 − 15.2) 2

χ2 = ∑ = + + +
E 37.2 22.8 24.8 15.2
230 Chapter Eleven

= 1.243 + 2.028 + 1.865 + 3.042 = 8.178

Step 5: Reject H0 since 8.178 > 6.635.
Conclude that the two drugs are not similar in curing the patients.

11.37 Step 1: H0: The distributions of responses from May 2009 and March 2000 are the same.
Η1: The distribution of responses from May 2009 and March 2000 are different.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(4 – 1) = 3
For α = .05, the critical value of χ2 is 7.815.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in the
table below.
Better More Laws &
Enforcement Restrictions Both Unsure Total
May 425 308 93 74 900
2009 (398.5) (319) (107.5) (75)
March 372 330 122 76 900
2000 (398.5) (319) (107.5) (75)
Total 797 638 215 150 1800

(O − E ) 2 (425 − 398.5) 2 (308 − 319) 2 (93 − 107.5) 2 (74 − 75) 2

χ2 =∑ = + + +
E 398.5 319 107.5 75

(372 − 398.5) 2 (330 − 319) 2 (122 − 107.5) 2 (76 − 75) 2

+ + + +
398.5 319 107.5 75
= 1.762 + .379 + 1.956 + .013 + 1.762 + .379 + 1.956 + .013 = 8.220
Step 5: Reject H0 since 8.220 > 7.815.
Conclude that the distribution of responses from May 2009 and March 2000 are different.

11.39 Step 1: H0: The distributions of opinions are homogeneous for the two groups of workers.
Η1: The distributions of opinions are not homogeneous for the two groups of workers.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(3 – 1) = 2
For α = .025, the critical value of χ2 is 7.378.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 231

Opinion
Favor Oppose Uncertain Total
Blue collar 44 39 12 95
Workers (42.59) (42.59) (9.83)
White collar 21 26 3 50
Workers (22.41) (22.41) (5.17)
Total 65 65 15 145

(O − E ) 2 (44 − 42.59) 2 (39 − 42.59) 2 (12 − 9.83) 2 (21 − 22.41) 2

χ2 =∑ = + + +
E 42.59 42.59 9.83 22.41

(26 − 22.41) 2 (3 − 5.17) 2

+ +
22.41 5.17
= .047 + .303 + .479 + .089 + .575 + .911 = 2.404
Step 5: Do not reject H0 since 2.404 < 7.378.
Conclude that the distributions of opinions are homogeneous for the two groups of workers.

Section 11.5

11.41 df = n – 1 = 25 – 1 = 24
a. χ2 for 24 df and .005 area in the right tail = 45.559
χ2 for 24 df and .995 area in the right tail = 9.886
The 99% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 ( 25 − 1)(35) ( 25 − 1)(35)
2
to = to = 18.4376 to 84.9686
χα / 2 χ 12−α / 2 45.559 9.886

b. χ2 for 24 df and .025 area in the right tail = 39.364

χ2 for 24 df and .975 area in the right tail = 12.401
The 95% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 ( 25 − 1)(35) ( 25 − 1)(35)
2
to = to = 21.3393 to 67.7365
χα / 2 χ 12−α / 2 39.364 12.401

c. χ2 for 24 df and .05 area in the right tail = 36.415

χ2 for 24 df and .95 area in the right tail = 13.848
The 90% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 ( 25 − 1)(35) ( 25 − 1)(35)
2
to = to = 23.0674 to 60.6586
χα / 2 χ 12−α / 2 36.415 13.848

As the confidence level decreases, the confidence interval for σ2 decreases in width.

11.43 a. H0: σ2 = 1.75, H1: σ2 > 1.75

232 Chapter Eleven

b. df = n –1 = 21 – 1 = 20
χ2 for 20 df and .025 area in the right tail = 34.170.

c. χ2 = (n – 1)s2/σ2 = (20 – 1)(1.97)/1.75 = 22.514

d. Do not reject H0 since 22.514 < 34.170.

11.45 a. H0: σ2 = 2.2, H1: σ2 ≠ 2.2

b. df = n – 1 = 18 – 1 = 17
χ2 for 17 df and .025 area in the right tail = 30.191
χ2 for 17 df and .975 area in the right tail = 7.564

c. χ2 = (n –1)s2/σ2 = (18 – 1)(4.6)/2.2 = 35.545

d. Reject H0 since 35.545 > 30.191.

11.47 a. df = n – 1 = 24 – 1 = 23
χ2 for 23 df and .01 area in the right tail = 41.638
χ2 for 23 df and .99 area in the right tail = 10.196
The 98% confidence interval for the population variance σ2 is

(n − 1) s 2 (n − 1) s 2 (24 − 1)(1.47) (24 − 1)(1.47)

2
to = to = .8120 to 3.3160
χα / 2 χ 12−α / 2 41.638 10.196

The 98% confidence interval for σ is .8120 to 3.3160 = .901 to 1.821

2 2
b. Step 1: H0: σ = 1.0, Hl: σ > 1.0
Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: For α = .01 with df = 23, the critical value of χ2 is 41.638.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 233

Step 4: χ2 = (n –1)s2/σ2 = (24 – 1)(1.47)/1.0 = 33.810

Step 5: Do not reject H0 since 33.810 < 41.638.
Conclude that the population variance is not greater than 1.0 square grams.

11.49 a. df = n – 1 = 25 – 1 = 24
χ2 for 24 df and .005 area in the right tail = 45.559
χ2 for 24 df and .995 area in the right tail = 9.886
The 99% confidence interval for σ 2 is

(n − 1) s 2 (n − 1) s 2 (25 − 1)(5200) (25 − 1)(5200)

2
to = to = 2739.3051 to 12,623.9126
χα / 2 χ 12−α / 2 45.559 9.886

The 99% confidence interval for σ is 2739.3051 to 12,623.9126 = 52.338 to 112.356

2 2
b. Step 1: H0: σ = 4200, Hl: σ ≠ 4200
Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: For α = .05 with df = 24, the critical values of χ2 are 12.401 and 39.364.
Step 4: χ2 = (n –1)s2/σ2 = (25 – 1)(5200)/4200 = 29.714
Step 5: Do not reject H0 since 12.401 < 29.714 < 39.364.
Conclude that the population variance is not different from 4200 square hours.

Supplementary Exercises

11.51 Step 1: H0: The percentage of people who consume total-bran cereal is the same for all four
brands.
Η1: The percentage of people who consume total-bran cereal is not the same for all four
brands.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 4, df = k – 1 = 4 – 1 = 3
For α = .05, the critical value of χ2 is 7.815.
Step 4:
Brand O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
A 212 .25 250 –38 1444 5.776
B 284 .25 250 34 1156 4.624
C 254 .25 250 4 16 .064
D 250 .25 250 0 0 .000
n = 1000 Sum = 10.464

χ2 = Σ (O – E)2/E = 10.464
Step 5: Reject H0 since 10.464 > 7.815.
234 Chapter Eleven

Conclude that the percentage of people who consume total-bran cereal is not the same for all four
brands.

11.53 Step 1: H0: The current distribution of results is the same as the distribution reported by
WSOC TV.
Η1: The current distribution of results differs from the distribution reported by WSOC
TV.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 5, df = k – 1 = 5 – 1 = 4
For α = .025, the critical value of χ2 is 11.143.
Step 4:
Song O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Blue Christmas 27 .0526 31.56 –4.56 20.794 .659
Carol of the Bells 117 .1545 92.70 24.30 590.490 6.370
Jingle Bell Rock 102 .1899 113.94 –11.94 142.564 1.251
White Christmas 268 .3991 239.46 28.54 814.532 3.402
Winter Wonderland 86 .2039 122.34 –36.34 1320.596 10.794
n = 600 Sum = 22.476

χ2 = Σ (O – E)2/E = 22.476
Step 5: Reject H0 since 22.476 > 11.143.
Conclude that the current distribution of results differs from the distribution reported by WSOC
TV.

11.55 Step 1: H0: The proportions of all allergic persons are the same over the four seasons.
Η1: The proportions of all allergic persons are not the same over the four seasons.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 4, df = k – 1 = 4 – 1 = 3
For α = .01, the critical value of χ2 is 11.345.
Step 4:
Season O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
Fall 18 .25 25 –7 49 1.960
Winter 13 .25 25 –12 144 5.760
Spring 31 .25 25 6 36 1.440
Summer 38 .25 25 13 169 6.760
n = 100 Sum = 15.920

χ2 = Σ (O – E)2/E = 15.920
Step 5: Reject H0 since 15.920 > 11.345.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 235

Conclude that the proportions of all allergic persons are not the same over the four seasons.

11.57 Step 1: H0: Political affiliation and concern about one’s own health care quality are
independent.
Η1: Political affiliation and concern about one’s own health care quality are dependent.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (3 – 1)(3 – 1) = 4
For α = .01, the critical value of χ2 is 13.277.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Democrat Independent Republican Total
Very 46 116 88 250
concerned (71.508) (124.860) (53.631)
Somewhat 100 143 61 304
concerned (86.954) (151.830) (65.216)
Not too/not at 110 188 43 341
all concerned (97.537) (170.309) (73.153)
Total 256 447 192 895

(O − E ) 2 (46 − 71.508) 2 (116 − 124.860) 2 (88 − 53.631) 2 (100 − 86.954) 2

χ2 =∑ = + + +
E 71.508 124.860 53.631 86.954

(143 − 151.830) 2 (61 − 65.216) 2 (110 − 97.537) 2 (188 − 170.309) 2 (43 − 73.153) 2
+ + + + +
151.830 65.216 97.537 170.309 73.153
= 9.099 + .629 + 22.025 + 1.957 + .514 + .273 + 1.592 + 1.838 + 12.429 = 50.356
Step 5: Reject H0 since 50.356 > 13.277.
Conclude that political affiliation and concern about one’s own health care quality are dependent.

11.59 Step 1: H0: Gender and marital status are not related for all persons who hold more than one
job.
Η1: Gender and marital status are related for all persons who hold more than one job.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(3 – 1) = 2
For α = .10, the critical value of χ2 is 4.605.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
236 Chapter Eleven

Gender Single Married Other Total

Male 72 209 39 320
(67.20) (199.04) (53.76)
Female 33 102 45 180
(37.80) (111.96) (30.24)
Total 105 311 84 500

(O − E ) 2 (72 − 67.20) 2 (209 − 199.04) 2 (39 − 53.76) 2 (33 − 37.80) 2

χ2 =∑ = + + +
E 67.20 199.04 53.76 37.80

(102 − 111.96) 2 (45 − 30.24) 2

+ +
111.96 30.24
= .343 + .498 + 4.052 + .610 + .886 + 7.204 = 13.593
Step 5: Reject H0 since 13.593 > 4.605.
Conclude that gender and marital status are related for all persons who hold more than one job.

11.61 Step 1: H0: The percentages of people with different opinions are similar for all four regions.
Η1: The percentages of people with different opinions are not similar for all four
regions.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (4 – 1)(3 – 1) = 6
For α = .01, the critical value of χ2 is 16.812.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Region Favor Oppose Uncertain Total
Northeast 56 33 11 100
(63.75) (29.75) (6.50)
Midwest 73 23 4 100
(63.75) (29.75) (6.50)
South 67 28 5 100
(63.75) (29.75) (6.50)
West 59 35 6 100
(63.75) (29.75) (6.50)
Total 255 119 26 400
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 237

(O − E ) 2 (56 − 63.75) 2 (33 − 29.75) 2 (11 − 6.50) 2 (73 − 63.75) 2

χ2 =∑ = + + +
E 63.75 29.75 6.50 63.75

(23 − 29.75) 2 (4 − 6.50) 2 (67 − 63.75) 2 (28 − 29.75) 2

+ + + +
29.75 6.50 63.75 29.75

(5 − 6.50) 2 (59 − 63.75) 2 (35 − 29.75) 2 (6 − 6.50) 2

+ + + +
6.50 63.75 29.75 6.50
= .942 + .355 + 3.115 + 1.342 + 1.532 + .962 + .166 + .103 + .346 + .354
+ .926 + .038 = 10.181
Step 5: Do not reject H0 since 10.181 < 16.812.
Conclude that the percentages of people with different opinions are similar for all four regions.

11.63 a. df = n – 1 = 10 – 1 = 9
χ2 for 9 df and .025 area in the right tail = 19.023
χ2 for 9 df and .975 area in the right tail = 2.700
The 95% confidence interval for the population variance σ2 is

(n − 1) s 2 (n − 1) s 2 (10 − 1)(7.2) (10 − 1)(7.2)

2
to = to = 3.4064 to 24.0000
χα / 2 χ 12−α / 2 19.023 2.700

The 95% confidence interval for σ is 3.4064 to 24.0000 = 1.846 to 4.899

b. df = n – 1 = 18 – 1 = 17
χ2 for 17 df and .025 area in the right tail = 30.191
χ2 for 17 df and .975 area in the right tail = 7.564
The 95% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 (18 − 1)(14.8) (18 − 1)(14.8)

2
to = to = 8.3336 to 33.2628
χα / 2 χ 12−α / 2 30.191 7.564

The 95% confidence interval for σ is 8.3336 to 33.2628 = 2.887 to 5.767

11.65 Step 1: H0: σ2 = 1.1, Hl: σ2 > 1.1

Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: df = n – 1 = 17 – 1 = 16
For α = .01, the critical value of χ2 is 28.845.
Step 4: χ2 = (n –1)s2/σ2 = (17– 1)(1.7)/1.1 = 24.727
Step 5: Do not reject H0 since 24.727 < 28.845.
Conclude that the population variance is not greater than 1.1.
238 Chapter Eleven

11.67 Step 1: H0: σ2 = 10.4, Hl: σ2 ≠ 10.4

Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: df = n – 1 = 18 – 1 = 17
For α = .05, the critical values of χ2 are 7.564 and 30.191.
Step 4: χ2 = (n –1)s2/σ2 = (18 – 1)(14.8)/10.4 = 24.192
Step 5: Do not reject H0 since 7.564 < 24.192 < 30.191.
Conclude that the population variance is not different from 10.4.

11.69 a. Step 1: H0: σ2 = 5000, Hl: σ2 < 5000

Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: df = n – 1 = 20 – 1 = 19
For α = .025, the critical value of χ2 is 8.907.
Step 4: χ2 = (n –1)s2/σ2 = (20 – 1)(3175)/5000 = 12.065
Step 5: Do not reject H0 since 12.065 > 8.907.
Conclude that the population variance is not less than 5000.
b. χ2 for 19 df and .01 area in the right tail = 36.191
χ2 for 19 df and .99 area in the right tail = 7.633
The 98% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 ( 20 − 1)(3175) ( 20 − 1)(3175)
2
to = to = 1666.8509 to 7903.1835
χα / 2 χ 12−α / 2 36.191 7.633

The 98% confidence interval for the population standard deviation σ is

1666.8509 to 7903.1835 = 40.827 to 88.900

11.71 a. df = n – 1 = 25 – 1 = 24
χ2 for 24 df and .005 area in the right tail = 45.559
χ2 for 24 df and .995 area in the right tail = 9.886
The 99% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 (25 − 1)(.19) (25 − 1)(.19)

2
to = to = .1001 to .4613
χα / 2 χ 12−α / 2 45.559 9.886

The 99% confidence interval for σ is .1001 to .4613 = .316 to .679

b. Step 1: H0: σ2 = .13, Hl: σ2 ≠ .13
Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: For α = .01 with df = 24, the critical values of χ2 are 9.886 and 45.559.
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 239

Step 4: χ2 = (n –1)s2/σ2 = (25 – 1)(.19)/.13 = 35.077

Step 5: Do not reject H0 since 9.886 < 35.077 < 45.559.
Conclude that the population variance is not different from .13.

11.73 a. ∑x = 4897, ∑x2 = 3,010,461

( x)
∑ x 2 − ∑n
2
(4897)2
3,010,461 −
s2 = = 8 = 1840.696429
n −1 8 −1
b. df = n – 1 = 8 – 1 = 7
χ2 for 7 df and .025 area in the right tail = 16.013
χ2 for 7 df and .975 area in the right tail = 1.690
The 95% confidence interval for σ2 is

(n − 1) s 2 (n − 1) s 2 (8 − 1)(1840.696429 ) (8 − 1)(1840 .696459 )

2
to = to = 804.6509 to
χα / 2 χ 12−α / 2 16.013 1.690

7624.1864

The 95% confidence interval for σ is 804.6509 to 7624.1864 = 28.37 to 87.32

2 2
c. Step 1: H0: σ = 750, Hl: σ ≠ 750
Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: For α = .05 with df = 7, the critical values of χ2 are 1.690 and 16.013.
Step 4: χ2 = (n –1)s2/σ2 = (8 – 1)(1840.696429)/750 = 17.180
Step 5: Reject H0 since 17.180 > 16.013.
Conclude that the population variance is different from 750 square dollars.

11.75 Step 1: H0: Opinions on disposal site are independent of gender.

Hl: Opinions on disposal site are dependent on gender.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(3 – 1) = 2
For α = .05, the critical value of χ2 is 5.991.
Step 4: From the given data we can calculate the following:
Total number opposed = 200(.60) = 120
Total number in favor = 200(.32) = 64
Total number undecided = 200(.08) = 16
Number of women opposed = 120(.65) = 78
Number of men opposed = 120 – 78 = 42
240 Chapter Eleven

Number of men in favor = 64(.625) = 40

Number of women in favor = 64 – 40 = 24
Number of women undecided = 110 – 78 – 24 = 8
Number of men undecided = 16 – 8 = 8
Using these results, we may construct the following contingency table of observations
and expected values.
Opposed In Favor Undecided Total
Women 78 24 8 110
(66.0) (35.2) (8.8)
Men 42 40 8 90
(54.0) (28.8) (7.2)
Total 120 64 16 200

(O − E ) 2 (78 − 66.0) 2 (24 − 35.2) 2 (8 − 8.8) 2 (42 − 54.0) 2

χ2 =∑ = + + +
E 66.0 35.2 8.8 54.0

(40 − 28.8) 2 (8 − 7.2) 2

+ +
28.8 7.2
= 2.182 + 3.564 + .073 + 2.667 + 4.356 + .089 = 12.931
Step 5: Reject H0 since 12.931 > 5.991.
Conclude that opinions on disposal site are dependent on gender.

11.77 Step 1: H0: The proportions of red and green marbles are the same in all five boxes.
Η1: The proportions of red and green marbles are not the same in all five boxes.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(5 – 1) = 4
For α = .05, the critical value of χ2 is 9.488.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in
the table below.
Box 1 2 3 4 5 Total
Red 20 14 23 30 18 105
(21) (21) (21) (21) (21)
Green 30 36 27 20 32 145
(29) (29) (29) (29) (29)
Total 50 50 50 50 50 250

(O − E ) 2 (20 − 21) 2 (14 − 21) 2 (23 − 21) 2 (30 − 21) 2 (18 − 21) 2
χ2 =∑ = + + + +
E 21 21 21 21 21
(30 − 29) 2 (36 − 29) 2 (27 − 29) 2 (20 − 29) 2 (32 − 29) 2
+ + + + +
29 29 29 29 29
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 241

= .048 + 2.333 + .190 + 3.857 + .429 + .034 + 1.690 + .138 + 2.793 + .310
= 11.822
Step 5: Reject H0 since 11.822 > 9.488.
Conclude that the proportions of red and green marbles are not the same in all five boxes.
11.79 Step 1: H0: A normal distribution is an appropriate model for these data.
Η1: A normal distribution is not an appropriate model for these data.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 10, df = k – 1 = 10 – 1 = 9
For α = .05, the critical value of χ2 is 16.919.
Step 4: We will find the probability values necessary to fill in the table used to calculate the chi-
square test statistic using the standard normal distribution table. For example, for the second row,
we will find p = P(−2 < z < −1.5) = P(z < −1.5) − P(z < −2) = .0668 − .0228 = .0440. Note the
symmetry in the standard normal corresponds to the symmetry in the probabilities, and, therefore,
in the expected counts.
Category O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
z score below −2 48 .0228 22.8 25.2 635.04 27.853
z score from −2 to −1.5 67 .0440 44.0 23.0 529.00 12.023
z score from −1.5 to −1 146 .0919 91.9 54.1 2926.81 31.848
z score from −1 to −0.5 248 .1498 149.8 98.2 9643.24 64.374
z score from −0.5 to 0 187 .1915 191.5 −4.5 20.25 .106
z score from 0 to 0.5 125 .1915 191.5 −66.5 4422.25 23.093
z score from 0.5 to 1 88 .1498 149.8 −61.8 3819.24 25.496
z score from 1 to 1.5 47 .0919 91.9 −44.9 2016.01 21.937
z score from 1.5 to 2 25 .0440 44.0 −19.0 361.00 8.205
z score of 2 or above 19 .0228 22.8 −3.8 14.44 .633
n = 1000 Sum =
215.568

χ2 = Σ (O – E)2/E = 215.568
Step 5: Reject H0 since 215.568 > 16.919.
Conclude that a normal distribution is not an appropriate model for these data.

11.81 a. Step 1: H0: All categories are equally likely.

Η1: All categories are not equally likely.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 4, df = k – 1 = 4 – 1 = 3
For α = .10, the largest reasonable significance level, the critical value of χ2 is 6.251.
Step 4:
242 Chapter Eleven

Category O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε

A 21 .25 25 −4 16 .64
B 26 .25 25 1 1 .04
C 31 .25 25 6 36 1.44
D 22 .25 25 −3 9 .36
n = 100 Sum = 2.48

χ2 = Σ (O – E)2/E = 2.48
Step 5: Do not reject H0 since 2.48 < 6.251.
Conclude that all categories are equally likely. Note that for any α < .10, χ2 > 6.251. Hence,
we would not reject H0 for any reasonable significance level.
b. We must have ∑ | O – E | = 14 and ∑(O – E) = 0. Under these constraints, the value of the chi-
square test statistic is maximized when the four values for O – E are 0, 0, 7, and –7. In such a

(0) 2 (0) 2 (7) 2 (7) 2

case, calculating the chi-square test statistic yields χ 2 = + + + = 3.92 .
25 25 25 25
Since 3.92 < 6.251, the p-value > .10.

Self-Review Test

1. b 2. a 3. c 4. a 5. b 6. b 7. c 8. b 9. a

10. Step 1: H0: The current distribution of race/ethnicity of obese and overweight people is the same as
the 2008 distribution.
Η1: The current distribution of race/ethnicity of obese and overweight people differs from
the 2008 distribution.
Step 2: Since this is a multinomial experiment, use the chi-square distribution.
Step 3: k = 6, df = k – 1 = 6 – 1 = 5
For α = .05, the critical value of χ2 is 11.070.
Step 4:
Race/ethnicity O p Ε = np Ο–Ε (Ο – Ε)2 (Ο – Ε) 2/Ε
African American 3048 .1562 3124 –76 5776 1.849
Am. Indian/Nat. Alaskan 137 .0073 146 –9 81 .555
Asian/Pac. Islander 645 .0286 572 73 5329 9.316
Hispanic/Latino 3169 .1526 3052 117 13,689 4.485
Other 18 .0006 12 6 36 3.000
White 12,983 .6547 13,094 –111 12,321 .941
n = 20,000 Sum = 20.146

χ2 = Σ (O – E)2/E = 20.146
 Introductory Statistics, Mann, Seventh Edition - Student’s Solutions Manual 243

Step 5: Reject H0 since 20.146 > 11.070.

Conclude that the current distribution of race/ethnicity of obese and overweight people differs from
the 2008 distribution.

11. Step 1: H0: Educational level and ever being divorced are independent.
Η1: Educational level and ever being divorced are dependent.
Step 2: Since this is a test of independence, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (2 – 1)(4 – 1) = 3
For α = .01, the critical value of χ2 is 11.345.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in the
table below.
Educational Level
Less than High school Some College
High school degree college degree Total
Divorced 173 158 95 53 479
(160.47) (136.04) (98.20) (84.30)
Never 162 126 110 123 521
Divorced (174.54) (147.96) (106.81) (91.70)
Total 335 284 205 176 1000

(O − E ) 2 (173 − 160.47) 2 (158 − 136.04) 2 (95 − 98.20) 2 (53 − 84.30) 2

χ2 =∑ = + + +
E 160.47 136.04 98.20 84.30

(162 − 174.54) 2 (126 − 147.96) 2 (110 − 106.81) 2 (123 − 91.70) 2

+ + + +
174.54 147.96 106.81 91.70
= .978 + 3.545 + .104 + 11.621 + .901 + 3.259 + .095 + 10.684 = 31.187
Step 5: Reject H0 since 31.187 > 11.345.
Conclude that educational level and ever being divorced are dependent.

12. Step 1: H0: The percentages of people who play the lottery often, sometimes, and never are the
same for each income group.
Η1: The percentages of people who play the lottery often, sometimes, and never are not the
same for each income group.
Step 2: Since this is a test of homogeneity, use the chi-square distribution.
Step 3: df = (R – 1)(C – 1) = (3 – 1)(3 – 1) = 4
For α = .05, the critical value of χ2 is 9.488.
Step 4: The expected frequencies are given in parentheses below the observed frequencies in the
table below.
244 Chapter Eleven

Income Group
Low Middle High Total
Play often 174 163 90 427
(170.80) (142.33) (113.87)
Play sometimes 286 217 120 623
(249.20) (207.67) (166.13)
Never Play 140 120 190 450
(180.00) (150.00) (120.00)
Total 600 500 400 1500

(O − E ) 2 (174 − 170.80) 2 (163 − 142.33) 2 (90 − 113.87) 2

χ2 =∑ = + +
E 170.80 142.33 113.87
(286 − 249.20) 2 (217 − 207.67) 2 (120 − 166.13) 2
+ + +
249.20 207.67 166.13
(140 − 180.00) 2 (120 − 150.00) 2 (190 − 120.00) 2
+ + +
180.00 150.00 120.00
= .060 + 3.002 + 5.004 + 5.434 + .419 + 12.809 + 8.889 + 6.000 + 40.833
= 82.450
Step 5: Reject H0 since 82.450 > 9.488.
Conclude that the percentages of people who play the lottery often, sometimes, and never are not the
same for each income group.

13. a. df = n – 1 = 20 – 1 = 19
χ2 for 19 df and .005 area in the right tail = 38.582
χ2 for 19 df and .995 area in the right tail = 6.844
The 99% confidence interval for the population variance σ2 is

(n − 1) s 2 (n − 1) s 2 (20 − 1)(.48) (20 − 1)(.48)

2
to = to = .2364 to 1.3326
χα / 2 χ 12−α / 2 38.582 6.844

The 99% confidence interval for σ is .2364 to 1.3326 = .486 to 1.154

b. Step 1: H0: σ2 = .25, Hl: σ2 > .25
Step 2: Since this is a test about a population variance, use the chi-square distribution.
Step 3: For α = .01 with df = 19, the critical value of χ2 is 36.191.
Step 4: χ2 = (n –1)s2/σ2 = (20 – 1)(.48)/.25 = 36.480
Step 5: Reject H0 since 36.480 > 36.191.
Conclude that the population variance exceeds .25 square ounce.