Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (15.74 MB, 756 trang )
Answers to Selected
Odd-Numbered Exercises
Chapter 1
1. a. Los Angeles Times, Oberlin Tribune, Gainesville Sun,
Washington Post
b. Duke Energy, Clorox, Seagate, Neiman Marcus
c. Vince Correa, Catherine Miller, Michael Cutler, Ken Lee
d. 2.97, 3.56, 2.20, 2.97
3. a. How likely is it that more than half of the sampled computers will need or have needed warranty service? What is
the expected number among the 100 that need warranty
service? How likely is it that the number needing warranty
service will exceed the expected number by more than 10?
b. Suppose that 15 of the 100 sampled needed warranty
service. How confident can we be that the proportion of all
such computers needing warranty service is between .08
and .22? Does the sample provide compelling evidence for
concluding that more than 10% of all such computers need
warranty service?
5. a. No. All students taking a large statistics course who participate in an SI program of this sort.
b. Randomization protects against various biases and helps
ensure that those in the SI group are as similar as possible to
the students in the control group.
c. There would be no firm basis for assessing the effectiveness of SI (nothing to which the SI scores could reasonably
be compared).
7. One could generate a simple random sample of all singlefamily homes in the city, or a stratified random sample by
taking a simple random sample from each of the 10 district
neighborhoods. From each of the selected homes, values of
all desired variables would be determined. This would be an
enumerative study because there exists a finite, identifiable
population of objects from which to sample.
9. a. Possibly measurement error, recording error, differences
in environmental conditions at the time of measurement, etc.
b. No. There is no sampling frame.
|
|
|
|
|
|
||
11. 6L 430
6H 769689
7L 42014202
7H
8L 011211410342
8H 9595578
9L 30
9H 58
The gap in the data—no scores in the high 70’s.
13. a. 12 2
leaf: ones digit
12 445
12 6667777
12 889999
13 00011111111
13 222222222233333333333333
13 44444444444444444455555555555555555555
13 6666666666667777777777
13 888888888888999999
14 0000001111
14 2333333
14 444
14 77
symmetry
b. Close to bell-shaped, center Ϸ 135, not insignificant dispersion, no gaps or outliers.
691
692
Answers to Selected Odd-Numbered Exercises
15. Crunchy
Creamy
2 2
644 3 0069
77220 4 00145
stem: tens
6320 5 003666
leaf: ones
222 6 258
55 7
0 8
Both sets of scores are rather spread out. There appear to be
no outliers. The distribution of crunchy scores appears to be
shifted to the right (toward larger values) of that for creamy
scores by something on the order of 10.
|
|
|
|
|
||
|
|
|
|
|
||
17. a. # Nonconforming
0
1
2
3
4
5
6
7
8
Frequency
Rel. freq.
7
12
13
14
6
3
3
1
1
60
.117
.200
.217
.233
.100
.050
.050
.017
.017
1.001
b. .917, .867, 1 Ϫ .867 ϭ .133
c. The histogram has a substantial positive skew. It is centered somewhere between 2 and 3 and spreads out quite a bit
about its center.
19. a. .99 (99%), .71 (71%)
b. .64 (64%), .44 (44%)
c. Strictly speaking, the histogram is not unimodal, but is
close to being so with a moderate positive skew. A much
larger sample size would likely give a smoother picture.
21. a. y
0
1
2
3
4
5
23. a.
b.
Freq.
Rel. freq.
17
.362
22
.468
6
.128
1
.021
0
.000
1
.021
47
1.000
.362, .638
Class
0–Ͻ100
100–Ͻ200
200–Ͻ300
300–Ͻ400
400–Ͻ500
500–Ͻ600
600–Ͻ700
700–Ͻ800
800–Ͻ900
Freq.
21
32
26
12
4
3
1
0
1
100
b.
z
0
1
2
3
4
5
6
7
8
Freq.
13
.277
11
.234
3
.064
7
.149
5
.106
3
.064
3
.064
0
.000
2
.043
47
1.001
.894, .830
Rel. freq.
.21
.32
.26
.12
.04
.03
.01
.00
.01
1.00
Rel. freq.
Class
Freq.
Rel. freq.
Density
0–Ͻ50
50–Ͻ100
100–Ͻ150
150–Ͻ200
200–Ͻ300
300–Ͻ400
400–Ͻ500
500–Ͻ600
600–Ͻ900
8
13
11
21
26
12
4
3
2
100
.08
.13
.11
.21
.26
.12
.04
.03
.02
1.00
.0016
.0026
.0022
.0042
.0026
.0012
.0004
.0003
.00007
c. .79
25.
Class
Freq.
Class
10–Ͻ20
20–Ͻ30
30–Ͻ40
40–Ͻ50
50–Ͻ60
60–Ͻ70
70–Ͻ80
8
14
8
4
3
2
1
40
1.1–Ͻ1.2
1.2–Ͻ1.3
1.3–Ͻ1.4
1.4–Ͻ1.5
1.5–Ͻ1.6
1.6–Ͻ1.7
1.7–Ͻ1.8
1.8–Ͻ1.9
Freq.
2
6
7
9
6
4
5
1
40
Original: positively skewed;
Transformed: much more symmetric, not far from bell-shaped.
27. a. The observation 50 falls on a class boundary.
b.
Class
Freq.
Rel. freq.
0–Ͻ50
50–Ͻ100
100–Ͻ150
150–Ͻ200
200–Ͻ300
300–Ͻ400
400–Ͻ500
500–Ͻ600
9
.18
19
.38
11
.22
4
.08
4
.08
2
.04
0
.00
1
.02
50
1.00
A representative (central) value is either a bit below or a bit
above 100, depending on how one measures center. There is a
great deal of variability in lifetimes, especially in values at the
upper end of the data. There are several candidates for outliers.
c.
Class
2.25–Ͻ2.75
2.75–Ͻ3.25
3.25–Ͻ3.75
3.75–Ͻ4.25
4.25–Ͻ4.75
4.75–Ͻ5.25
5.25–Ͻ5.75
5.75–Ͻ6.25
Freq.
Rel. freq.
2
.04
2
.04
3
.06
8
.16
18
.36
10
.20
4
.08
3
.06
50
1.00
There is much more symmetry in the distribution of the
ln(x) values than in the x values themselves, and less variability. There are no longer gaps or obvious outliers.
d. .38, .14
Answers to Selected Odd-Numbered Exercises
29. Complaint
J
F
B
M
C
N
O
31.
Freq.
Rel. freq.
10
9
7
4
3
6
21
60
.1667
.1500
.1167
.0667
.0500
.1000
.3500
1.0001
693
59. a. ED: .4, .10, 2.75, 2.65;
Non-Ed: 1.60, .30, 7.90, 7.60
b. ED: 8.9 and 9.2 are mild outliers, and 11.7 and 21.0 are
extreme outliers.
There are not outliers in the non-ED sample.
c. Four outliers for ED, none for non-ED. Substantial positive skewness in both samples; less variability in ED
(smaller fs), and non-ED observations tend to be somewhat
larger than ED observations.
Class
Freq.
Cum. freq.
Cum. rel. freq.
0–Ͻ4
4–Ͻ8
8–Ͻ12
12–Ͻ16
16–Ͻ20
20–Ͻ24
24–Ͻ28
2
14
11
8
4
0
1
2
16
27
35
39
39
40
.050
.400
.675
.875
.975
.975
1.000
33. a. ෆx ϭ 192.57, ~
x ϭ 189.0
b. New ෆx ϭ 189.71; ~
x unchanged
c. 191.0, 7.14%
d. 122.6
~
35. a. xෆ ϭ 12.55, x ϭ 12.50, ෆxtr(12.5) ϭ 12.40. Deletion of the
largest observation (18.0) causes ~
x and ෆxtr to be a bit smaller
than ෆx.
b. By at most 4.0
c. No; multiply the values of xෆ and ~
x
by the conversion factor 1/2.2.
37. ෆxtr(10) ϭ 11.46
x ϭ 1.009
b. .383
39. a. ෆx ϭ 1.0297, ~
41. a. .7
b. Also .7
c. 13
43. ~
x ϭ 68.0, ෆxtr(20) ϭ 66.2, ෆxtr(30) ϭ 67.5
45. a. ෆx ϭ 115.58; the deviations are .82, .32, Ϫ.98, Ϫ.38, .22
b. .482, .694
c. .482
d. .482
47. ෆx ϭ 116.2, s ϭ 25.75. The magnitude of s indicates a substantial amount of variation about the center (a “representative” deviation of roughly 25).
61. Outliers, both mild and extreme, only at 6 A.M. Distributions
at other times are quite symmetric. Variability increases
somewhat until 2 P.M. and then decreases slightly, and the
same is true of “typical” gasoline-vapor coefficient values.
63.
6
34
7
17
8
4589
9
1
10
12667789
11
122499
12
2
13
1
~
ෆx ϭ 9.96, x ϭ 10.6, s ϭ 1.7594, fs ϭ 2.3, no outliers, negative skew
65. a. Representative value ϭ 90. Reasonably symmetric, unimodal, somewhat bell-shaped, fair amount of variability.
b. .9231, .9053
c. .48
67. a. M: ෆx ϭ 3.64, ~
x ϭ 3.70, s ϭ .269, fs ϭ .40
F: ෆx ϭ 3.28, ~
x ϭ 3.15, s ϭ .478, fs ϭ .50
Female values are typically somewhat smaller than male
values, and show somewhat more variability. An M boxplot
shows negative skew whereas an F boxplot shows positive
skew.
b. F: xෆtr(10) ϭ 3.24 M: xෆtr(10) ϭ 3.652 Ϸ 3.65
b. 189.14, 1.87
69. a. ෆy ϭ axෆ ϩ b, sy2 ϭ a2sx2
55. a. 33
b. No
c. Slight positive skewness in the middle half, but rather
symmetric overall. The extent of variability appears substantial.
d. At most 32
71. a. The mean, median, and trimmed mean are virtually identical, suggesting a substantial amount of symmetry in the
data; the fact that the quartiles are roughly the same distance
from the median and that the smallest and largest observations are roughly equidistant from the center provides additional support for symmetry. The standard deviation is quite
small relative to the mean and median.
b. See the comments of (a). In addition, using 1.5(Q3 Ϫ Q1)
as a yardstick, the two largest and three smallest observations are mild outliers.
73. ෆx ϭ .9255, s ϭ .0809, ~
x ϭ .93, small amount of variability,
slight bit of skewness
75. a. The “five-number summaries” ( ~
x, the two fourths, and
57. a. Yes. 125.8 is an extreme outlier and 250.2 is a mild
outlier.
b. In addition to the presence of outliers, there is positive
skewness both in the middle 50% of the data and, excepting
the outliers, overall. Except for the two outliers, there appears
to be a relatively small amount of variability in the data.
the smallest and largest observations) are identical and
there are no outliers, so the three individual boxplots are
identical.
b. Differences in variability, nature of gaps, and existence
of clusters for the three samples.
c. No. Detail is lost.
49. a. 56.80, 197.8040
b. .5016, .708
51. a. 1264.766, 35.564
b. .351, .593
53. a. 2.74, 3.88
d. At most .40
b. 1.14
e. 1.19
c. Unchanged
694
77. a.
Answers to Selected Odd-Numbered Exercises
0
1
2
3
4
5
6
7
8
9
10
HI
|
|
|
|
|
|
|
|
|
|
||
2355566777888
0000135555
00257
0033
0057
stem: ones
044
leaf: tenths
05
8
0
3
22.0, 24.5
b.
Class
Freq.
Rel. freq.
Density
0–Ͻ2
2–Ͻ4
4–Ͻ6
6–Ͻ10
10–Ͻ20
20–Ͻ30
23
9
7
4
1
2
.500
.196
.152
.087
.022
.043
.250
.098
.076
.022
.002
.004
79. a. ෆxnϩ1 ϭ (nxෆn ϩ xnϩ1)/(n ϩ 1)
c. 12.53, .532
81. A substantial positive skew (assuming unimodality)
83. a. All points fall on a 45° line. Points fall below a 45° line.
b. Points fall well below a 45° line, indicating a substantial
positive skew.
Chapter 2
1. a. S ϭ {1324, 3124, 1342, 3142, 1423, 1432, 4123, 4132,
2314, 2341, 3214, 3241, 2413, 2431, 4213, 4231}
b. A ϭ {1324, 1342, 1423, 1432}
c. B ϭ {2314, 2341, 3214, 3241, 2413, 2431, 4213, 4231}
d. A ʜ B ϭ {1324, 1342, 1423, 1432, 2314, 2341, 3214,
3241, 2413, 2431, 4213, 4231},
A ʝ B contains no outcomes (A and B are disjoint),
AЈ ϭ {3124, 3142, 4123, 4132, 2314, 2341, 3214, 3241,
2413, 2431, 4213, 4231}
3. a. A ϭ {SSF, SFS, FSS}
b. B ϭ {SSF, SFS, FSS, SSS}
c. C ϭ {SFS, SSF, SSS}
d. CЈ ϭ {FFF, FSF, FFS, FSS, SFF},
A ʜ C ϭ {SSF, SFS, FSS, SSS},
A ʝ C ϭ {SSF, SFS},
B ʜ C ϭ {SSF, SFS, FSS, SSS} ϭ B,
B ʝ C ϭ {SSF, SFS, SSS} ϭ C
5. a. S ϭ {(1, 1, 1), (1, 1, 2), (1, 1, 3), (1, 2, 1), (1, 2, 2), (1, 2, 3),
(1, 3, 1), (1, 3, 2), (1, 3, 3), (2, 1, 1), (2, 1, 2), (2, 1, 3),
(2, 2, 1), (2, 2, 2), (2, 2, 3), (2, 3, 1), (2, 3, 2), (2, 3, 3), (3, 1, 1),
(3, 1, 2), (3, 1, 3), (3, 2, 1), (3, 2, 2), (3, 2, 3), (3, 3, 1),
(3, 3, 2), (3, 3, 3)}
b. {(1, 1, 1), (2, 2, 2), (3, 3, 3)}
c. {(1, 2, 3), (1, 3, 2), (2, 1, 3), (2, 3, 1), (3, 1, 2), (3, 2, 1)}
d. {(1, 1, 1), (1, 1, 3), (1, 3, 1), (1, 3, 3), (3, 1, 1), (3, 1, 3),
(3, 3, 1), (3, 3, 3)}
7. a. There are 35 outcomes in S.
b. {AABABAB,
AABAABB, AAABBAB, AAABABB, AAAABBB}
11. a. .07
b. .30
c. .57
13. a. .36
d. .47
b. .64
e. .17
c. .53
f. .75
15. a. .572
b. .879
17. a. There are statistical software packages other than SPSS
and SAS.
b. .70
c. .80
d. .20
19. a. .8841
b. .0435
21. a. .10
e. .31
b. .18, .19
f. .69
23. a. .067
b. .400
25. a. .98
27. a. .1
c. .41
c. .933
b. .02
b. .7
d. .59
c. .03
d. .533
d. .24
c. .6
29. a. 676; 1296
b. 17,576; 46,656
1,679,616
d. .942
31. a. 243
b. 3645 days (roughly 10 yr)
33. a. 362,880
35. a. .0048
37. a. 60
39. a. .0839
41. a. .929
c. 456,976;
b. 131,681,894,400
b. .0054
b. 10
c. .9946
c. 2100
d. .2885
c. .0456
b. .24975
b. .0714
c. .99997520
43. .000394, .00394, .00001539
45. a. .447, .500, .200
47. a. .50
d. .375
b. .50
e. .769
b. .400, .447
c. .211
c. .625
49. .217, .178
51. .436, .581
53. .083
55. .236
59. a. .21
b. .455
61. a. .578, .278, .144
63. b. .54
c. .68
c. .264, .274
b. 0, .457, .543
d. .74
e. .7941
65. P(Mean⏐S) ϭ .3922, P(Median⏐S) ϭ .2941, so Mean and
Median are most and least likely, respectively.
67. .000329; very uneasy.
Answers to Selected Odd-Numbered Exercises
69. a. .126
e. .5325
b. .05
c. .1125
f. .2113
71. a. .300
b. .820
d. .2725
93. .45, .32
95. a. .0083
c. .146
b. .2
99. a. .956
b. .00421
b. .994
101. .926
79. .0059
103. a. .018
81. a. .95
105. a. .883, .117
83. a. .10, .20
b. 0
b. .601
b. 23
c. .156
107. 1 Ϫ (1 Ϫ p1)(1 Ϫ p2) и и и и и (1 Ϫ pn)
c. (1 Ϫ p)3
85. a. p(2 Ϫ p)
b. 1 Ϫ (1 Ϫ p)n
3
d. .9 ϩ (1 Ϫ p) (.1)
e. .1(1 Ϫ p)3/[.9 ϩ .1(1 Ϫ p)3] ϭ .0137 for p ϭ .5
109. a. .0417
b. .375
111. P(hire #1) ϭ 6/24 for s ϭ 0, ϭ 11/24 for s ϭ 1,
ϭ 10/24 for s ϭ 2, and ϭ 6/24 for s ϭ 3, so s ϭ 1 is best.
87. .8588, .9897
113. 1/4 ϭ P(A1 ʝ A2 ʝ A3)
P(A1) и P(A2) и P(A3) ϭ 1/8
89. [2(1 Ϫ )]/(1 Ϫ 2)
91. a. .333, .444
c. .2
97. .905
75. .401, .722
77. a. .06235
695
b. .150
c. .291
Chapter 3
1. x ϭ 0 for FFF; x ϭ 1 for SFF, FSF, and FFS; x ϭ 2 for SSF,
SFS, and FSS; and x ϭ 3 for SSS
3. Z ϭ average of the two numbers, with possible values 2/2,
3/2, . . . , 12/2; W ϭ absolute value of the difference, with
possible values 0, 1, 2, 3, 4, 5
5. No. In Example 3.4, let Y ϭ 1 if at most three batteries are
examined and let Y ϭ 0 otherwise. Then Y has only two
values.
7. a. {0, 1, . . . , 12}; discrete
c. {1, 2, 3, . . . }; discrete
e. {0, c, 2c, . . . , 10,000c}, where c is the royalty per book;
discrete g. {xϺ m Յ x Յ M} where m (M) is the minimum (maximum) possible tension; continuous
9. a. {2, 4, 6, 8, . . . }, that is, {2(1), 2(2), 2(3), 2(4), . . . }, an
infinite sequence; discrete
b. {2, 3, 4, 5, 6, . . . }, that is, {1 ϩ 1, 1 ϩ 2, 1 ϩ 3, 1 ϩ
4, . . . }, an infinite sequence; discrete
11. a. p(4) ϭ .45, p(6) ϭ .40, p(8) ϭ .15, p(x) ϭ 0 for x
6, or 8
c. .55, .15
13. a. .70
d. .71
b. .45
e. .65
4,
c. .55
f. .45
15. a. (1, 2), (1, 3), (1, 4), (1, 5), (2, 3), (2, 4), (2, 5), (3, 4),
(3, 5), (4, 5)
b. p(0) ϭ .3, p(1) ϭ .6, p(2) ϭ .1
c. F(x) ϭ 0 for x Ͻ 0, ϭ .3 for 0 Յ x Ͻ 1, ϭ .9 for 1 Յ
x Ͻ 2, and ϭ 1 for 2 Յ x
17. a. .81
b. .162
UUUAA; .00405
c. It is A; AUUUA, UAUUA, UUAUA,
b. F(x) ϭ 0 for x Ͻ 1, ϭ .477 for 1 Յ x Ͻ 2, ϭ .602 for
2 Յ x Ͻ 3, . . . , ϭ .954 for 8 Յ x Ͻ 9, ϭ 1 for x Ն 9
c. .602, .301
23. a. .20
b. .33
25. a. p(y) ϭ (1 Ϫ p)y и p
c. .78
d. .53
for y ϭ 0, 1, 2, 3, . . .
27. a. 1234, 1243, 1324, . . . , 4321
b. p(0) ϭ 9/24, p(1) ϭ 8/24, p(2) ϭ 6/24, p(3) ϭ 0,
p(4) ϭ 1/24
29. a. 2.06
b. .9364
c. .9677
d. .9364
31. .74, .8602, .85
33. a. p
b. p(1 Ϫ p)
c. p
35. E[h3(X)] ϭ 2.4667, E[h4(X)] ϭ 2.667, so 4 copies is better.
37. E(X ) ϭ (n ϩ 1)/2, E(X 2) ϭ (n ϩ 1)(2n ϩ 1)/6, V(X ) ϭ
(n2 Ϫ 1)/12
39. 2.3, .81, 88.5, 20.25
43. E(X Ϫ c) ϭ E(X) Ϫ c, E(X Ϫ ) ϭ 0
47. a. .515
f. .000
b. .218
g. .595
49. a. .354
b. .115
c. .011
d. .480
e. .965
c. .918
51. a. 6.25
b. 2.17
c. .030
53. a. .403
b. .787
c. .774
55. .1478
57. .407, independence
19. p(0) ϭ .09, p(1) ϭ .40, p(2) ϭ .32, p(3) ϭ .19
59. a. .017
21. a. p(x) ϭ .301, .176, .125, .097, .079, .067, .058, .051, .046
for x ϭ 1, 2, . . . , 9
61. When p ϭ .9, the probability is .99 for A and .9963 for B. If
p ϭ .5, these probabilities are .75 and .6875, respectively.
b. .811, .425
c. .006, .902, .586
696
Answers to Selected Odd-Numbered Exercises
63. The tabulation for p Ͼ .5 is unnecessary.
89. a. 4
65. a. 20, 16
91. a. .221
b. 70, 21
67. P(⏐X Ϫ ⏐Ն 2) ϭ .042 when p ϭ .5 and ϭ .065 when
p ϭ .75, compared to the upper bound of .25. Using k ϭ 3 in
place of k ϭ 2, these probabilities are .002 and .004, respectively, whereas the upper bound is .11.
69. a. .114
b. .879
c. .121
distribution with n ϭ 15, p ϭ .10
d. Use the binomial
75. a. nb(x; 2, .5)
b. .033
b. .188
c. .688
d. 2, 4
77. nb(x; 6, .5), 6
79. a. .932
e. .251
b. .065
81. a. .011
c. .068
b. .441
83. Poisson(5)
85. a. .122, .809, .283
c. .530, .011
87. a. .099
b. .135
d. .492
c. .554, .459
a. .492
b. 6,800,000
c. p(x; 20.106)
95. b. 3.114, .405, .636
97. a. b(x; 15, .75)
b. .686
c. .313
d. 11.25, 2.81
e. .310
99. .991
101. a. p(x; 2.5)
b. .067
c. .109
103. 1.813, 3.05
71. a. h(x; 15, 10, 20) for x ϭ 5, . . . , 10
b. .0325
c. .697
73. a. h(x; 10, 10, 20)
c. At least Ϫln(.1)/2 Ϸ 1.1513 years
b. .215
d. .945
105. p(2) ϭ p2, p(3) ϭ (1 Ϫ p)p2, p(4) ϭ (1 Ϫ p)p2, p(x) ϭ
[1 Ϫ p(2) Ϫ . . . Ϫ p(x Ϫ 3)](1 Ϫ p)p2 for x ϭ 5, 6, 7, . . . ;
.99950841
107. a. .0029
109. a. .135
b. .0767, .9702
b. .00144
c.
Α x∞ϭ0[p(x; 2)]5
111. 3.590
113. a. No
b. .0273
115. b. .6p(x; ) ϩ .4p(x; )
c. ( ϩ )/2
d. ( Ϫ )2/4 ϩ ( ϩ )/2
b. .133
b. 12, 3.464
117.
Α i10ϭ1(piϩjϩ1 ϩ piϪjϪ1)pi, where pk ϭ 0 if k Ͻ 0 or k Ͼ 10.
121. a. 2.50
c. 2
b. 3.1
Chapter 4
1. a. .25
b. .50
3. b. .5
c. .6875
25. b. 1.8(90th percentile for X) ϩ 32
c. a(X percentile) ϩ b
c. .4375
d. .6328
d. .578
27. 0, 1.814
7. a. f(x) ϭ .1 for 25 Յ x Յ 35 and 0 otherwise
b. .20
c. .40
d. .20
29. a. 2.14
d. .97
9. a. .562
31. a. 2.54
5. a. .375
b. .125
c. .297
b. .438, .438
c. .071
11. a. .25
b. .1875
c. .9375
e. f(x) ϭ x/2 for 0 Ͻ x Ͻ 2
f. 1.33
g. .222, .471
h. 2
d. 1.4142
13. a. 3
b. 0 for x Յ 1, 1 Ϫ x᎐3 for x Ͼ 1
c. .125, .088
d. 1.5, .866
e. .924
x9
x 10
c. 1.17
c. Ϫ.42
b. 1.34
33. a. .9918
b. .0082
35. a. .3336
c. .5795
b. Approximately 0
d. 6.524
e. .8028
37. a. 0, .5793, .5793
39. a. 36.7
15. a. F(x) ϭ 0 for x Յ 0, ϭ 90[ᎏ9ᎏ Ϫ ᎏ1ᎏ0 ] for 0 Ͻ x Ͻ 1, ϭ 1 for
xՆ1
b. .0107
c. .0107, .0107
d. .9036
e. .818, .111
f. .3137
41. .002
17. a. A ϩ (B Ϫ A)p
b. E(X) ϭ (A ϩ B)/2, ⌾ ϭ
(B Ϫ A)/͙12
ෆ
c. [Bn ϩ 1 Ϫ An ϩ 1]/[(n ϩ 1)(B Ϫ A)]
45. 7.3%
19. a. .597
b. .369
c. f(x) ϭ .3466 Ϫ .25 ln(x) for 0 Ͻ x Ͻ 4
b. .81
e. 2.41
c. .8664
c. Ͻ 87.6 or Ͼ 120.4
b. .3174, no
b. 22.225
c. 3.179
43. 10, .2
47. 21.155
49. a. .1190, .6969
b. .0021
c. .7054
21. 314.79
d. Ͼ5020 or Ͻ1844 (using z.0005 ϭ 3.295)
23. 248, 3.60
e. Normal, ϭ 7.576, ϭ 1.064, .7054
Answers to Selected Odd-Numbered Exercises
697
51. .3174 for k ϭ 1, .0456 for k ϭ 2, .0026 for k ϭ 3, as compared to the bounds of 1, .25, and .111, respectively.
97. There is substantial curvature in the plot. is a scale
parameter (as is for the normal family).
53. a. Exact: .212, .577, .573; Approximate: .211, .567, .596
b. Exact: .885, .575, .017; Approximate: .885, .579, .012
c. Exact: .002, .029, .617; Approximate: .003, .033, .599
99. a. F(y) ϭ ᎏ48ᎏ (y 2 Ϫ y 3/18) for 0 Յ y Յ 12
b. .259, .5, .241
c. 6, 43.2, 7.2
d. .518
e. 3.75
55. a. .9409
101. a. f(x) ϭ x2 for 0 Յ x Ͻ 1 and ϭ ᎏ4ᎏ Ϫ ᎏ4ᎏx for 1 Յ x Յ ᎏ3ᎏ
b. .917
c. 1.213
7
b. .9943
57. b. Normal, ϭ 239, 2 ϭ 12.96
59. a. 1
b. 1
1
c. .982
61. a. .449, .699, .148
d. .129
b. .05, .018
63. a. short d plan #1 better, whereas long d plan #2 better
b. 1\ ϭ 10 d E[h1 (X)] ϭ 100, E[h2(X)] ϭ 112.53
1\ ϭ 15 d E[h1 (X)] ϭ 150, E[h2(X)] ϭ 138.51
65. a. .238
f. .713
b. .238
c. .313
67. a. .424
~ Ͻ 24
b. .567,
d. .653
c. 60
e. .653
d. 66
103. a. .9162
b. .9549
c. 1.3374
b. .0663
c. (72.97, 119.03)
107. b. F(x) ϭ 0 for x Ͻ Ϫ1, ϭ (4x Ϫ x3/3)/9 ϩ
x Յ 2, and ϭ 1 for x Ͼ 2
~Ͼ0
c. No. F(0) Ͻ .5 d
5
d. Y ϳ Bin(10, ᎏ2ᎏ7 )
73. a. .826, .826, .0636
b. .664
c. 172.727
77. a. 123.97, 117.373
b. .5517
c. .1587
113. b.
c.
d.
e.
b. .9573
f. 125.90
c. .0414
83. ␣ ϭ 
85. b. [⌫(␣ ϩ ) и ⌫(m ϩ )]/[⌫(␣ ϩ  ϩ m) и ⌫()], /(␣ ϩ )
87. Yes, since the pattern in the plot is quite linear.
89. Yes
91. Yes
93. Plot ln(x) vs. z percentile. The pattern is straight, so a lognormal population distribution is plausible.
11
ᎏᎏ
27
for Ϫ1 Յ
109. a. .368, .828, .460
b. 352.53
c. 1/ и exp[Ϫexp(Ϫ(x Ϫ ␣)/)] и exp(Ϫ(x Ϫ ␣)/)
~ ϭ 182.99
d. ␣
e. ϭ 201.95, mode ϭ 150,
111. a.
b. No
c. 0
d. (␣ Ϫ 1)
e. Ϫ 2
81. a. 149.157, 223.595
d. 148.41
e. 9.57
7
105. a. .3859
69. a. ʝ Ai
b. Exponential with ϭ .05
c. Exponential with parameter n
79. a. 68.0, 122.1
b. .3204
c. .7257, skewness
3
p(1 Ϫ exp(Ϫ1x)) ϩ (1 Ϫ p)(1 Ϫ exp(Ϫ2x)) for x Ն 0
p/1 ϩ (1 Ϫ p)/2
V(X) ϭ 2p/12 ϩ 2(1 Ϫ p)/22 Ϫ 2
1, CV Ͼ 1
f. CV Ͻ 1
115. a. Lognormal
b. 1
c. 2.72, .0185
119. a. Exponential with ϭ 1
c. Gamma with parameters ␣ and c
121. a. (1/365)3
b. (1/365)2
c. .000002145
123. b. Let u1, u2, u3, . . . be a sequence of observations from a
Unif[0, 1] distribution (a sequence of random numbers).
Then with xi ϭ (Ϫ.1)ln(1 Ϫ ui), the xi’s are observations
from an exponential distribution with ϭ 10.
125. g(E(X)) Յ E(g(X))
127. a. 710, 84.423, .684
b. .376
95. The pattern in the plot is quite linear; it is very plausible that
strength is normally distributed.
Chapter 5
1. a. .20
b. .42
c. At least one hose is in use at each
pump; .70.
d. pX (x) ϭ .16, .34, .50 for x ϭ 0, 1, 2,
respectively; pY (y) ϭ .24, .38, .38 for y ϭ 0, 1, 2, respectively; .50
e. No; p(0, 0) pX (0) и pY (0)
b. eϪϪ и [1 ϩ ϩ ]
11. a. eϪϪ и x и y/x!y!
Ϫ(ϩ)
m
и ( ϩ ) /m!; Poisson ( ϩ )
c. e
3. a. .15
15. a. F(y) ϭ 1 Ϫ eϪy ϩ (1 Ϫ eϪy)2 Ϫ (1 Ϫ eϪy)3 for y Ն 0
b. 2/3
b. .40
5. a. .054
b. .00018
7. a. .030
d. .380
b. .120
e. Yes
c. .22
d. .17, .46
c. .300
9. a. 3/380,000
b. .3024
c. .3593
d. 10Kx2 ϩ .05 for 20 Յ x Յ 30
e. No
13. a. eϪxϪy for x Ն 0, y Ն 0
d. .330
b. .400
c. .594
17. a. .25
b. .318
c. .637
ෆ2ෆϪ
ෆෆx2ෆ/R2 for ϪR Յ x Յ R; no
d. fX(x) ϭ 2͙R
19. a. K(x2 ϩ y2)/(10Kx2 ϩ .05); K(x2 ϩ y2)/(10Ky2 ϩ .05)
b. .556, .549
c. 25.37, 2.87
698
Answers to Selected Odd-Numbered Exercises
21. a. f(x1, x2, x3)/f X1 ,X2 (x1, x2)
b. f(x1, x2, x3)/f X1(x1)
23. .15
59. a. .9986, .9986
b. .9015, .3970
c. .8357
d. .9525, .0003
25. L2
61. a. 3.5, 2.27, 1.51
27. .25 hr
63. a. .695
b. 4.0675 Ͼ 2.6775
65. a. .9232
2
29. Ϫᎏ3ᎏ
31. a. Ϫ.1082
37. a. xෆ
p(xෆ )
b. s2
p(s2)
|
0
112.5
| .38
312.5 800
.20
|
Probability
.30
.1
.2
.3
.4
.5
| .000
.000
.000
.001
.005
.027
|
.6
| .088
Probability
|
|
.12
, E(S 2) ϭ 212.25 ϭ 2
0
Proportion
p(xෆ )
b. .85
, E(X
ෆ) ϭ ϭ 44.5
.04 .20 .25 .12 .30 .09
39. Proportion
41. a. xෆ
69. a. 2400
25 32.5 40 45 52.5 65
|
.7
.8
.9
1.0
.201
.302
.269
.107
73. a. Approximately normal with mean ϭ 105, SD ϭ 1.2649;
Approximately normal with mean ϭ 100, SD ϭ 1.0142
b. Approximately normal with mean ϭ 5, SD ϭ 1.6213
c. .0068
d. .0010, yes
75. a. .2, .5, .3 for x ϭ 12, 15, 20; .10, .35, .55 for y ϭ 12, 15, 20
b. .25
c. No
d. 33.35
e. 3.85
.01
2
3
79. Ϸ 1
.30
.40
.22
.08
2.5
.16
.24
.25
.20
|
b. .9788
1
2
p(r)
71. a. 158, 430.25
c. 2400, 41.77
0
1.5
|
b. 1205; independence
77. a. 3/81,250
b. fX (x) ϭ k(250x Ϫ 10x2) for 0 Յ x Յ 20
1
and ϭ k(450x Ϫ 30x2 ϩ ᎏ2ᎏx3) for 20 Ͻ x Յ 30; fY (y) results
from substituting y for x in fX (x). They are not independent.
c. .355
d. 25.969
e. 204.6154, Ϫ.894
f. 7.66
1
c. r
b. .9660
67. .1588
b. Ϫ.0131
|
b. 15.4, 75.94, 8.71
3
.10
3.5
4
.04
81. a. 400 min
47. a. .6826
b. .1056
83. 97
49. a. .6026
b. .2981
85. .9973
b. 70
89. b, c. Chi-squared with ϭ n.
51. .7720
53. a. .0062
b. 0
91. a. 2W /( 2W ϩ 2E )
55. a. .9838
b. .8926
93. 26, 1.64
95. a. .6
57. .9616
b. .9999
ෆ
b. U ϭ X ϩ ͙1
Ϫ 2 Y
Chapter 6
~
1. a. 8.14, X
b. .77, X
ෆ
d. .148
e. .204, S/X
ෆ
15. a. ˆ ϭ ΑX 2i /2n
c. 1.66, S
17. b. .444
3. a. 1.348, X
b. 1.348, X
ෆ
ෆ
c. 1.781, X
ϩ
1.28S
ෆ
d. .6736
e. .0905
5. N xෆ ϭ 1,703,000;
1,601,438.281
7. a. 120.6
΄
΅
19. a. pˆ ϭ 2ˆ Ϫ .30 ϭ .20
b. pˆ ϭ (100ˆ Ϫ 9)/70
21. b. ␣ˆ ϭ 5, ˆ ϭ 28.0/⌫(1.2)
T Ϫ N dෆ ϭ 1,591,300;
b. 1,206,000
9. a. 2.11
b. .119
p1q1
p q 1/2
11. b. ᎏ ᎏ ϩ ᎏ2 ᎏ2
n1
n2
c. .80
T и (xෆ / ෆy) ϭ
d. 120.0
c. Use pˆ i ϭ xi/ni and qˆ i ϭ 1 Ϫ pˆ i
23. ˆ 1 ϭ xෆ, ˆ 2 ϭ yෆ, estimate of (1 Ϫ 2) is ෆx Ϫ ෆy.
25. a. 384.4, 18.86
b. 415.42
29. a. ˆ ϭ min(Xi), ˆ ϭ n/Α[Xi Ϫ min(Xi)]
b. .64, .202
33. With xi ϭ time between birth i Ϫ 1 and birth i, ˆ ϭ
6/Α6i ϭ1 ixi ϭ .0436.
in place of pi and qi in part (b) for i ϭ 1, 2.
35. 29.5
d. Ϫ.245
37. 1.0132
e. .041
b. 74.505
Answers to Selected Odd-Numbered Exercises
699
Chapter 7
1. a. 99.5%
b. 85%
3. a. Narrower
c. 2.96
b. No
5. a. (4.52, 5.18)
c. .55
d. 94
c. No
d. 1.15
d. No
35. a. 95% CI: (23.1, 26.9)
b. 95% PI: (17.2, 32.8), roughly 4 times as wide
37. a. (.888, .964)
c. (.634, 1.218)
b. (4.12, 5.00)
7. By a factor of 4; the width is decreased by a factor of 5.
9. a. (xෆ Ϫ 1.645/͙n
ෆ, ∞); (4.57, ∞)
ෆ, ∞)
c. (Ϫ∞, ෆx ϩ z␣ и /͙n
ෆ);
b. (xෆ Ϫ z␣ и /͙n
(Ϫ∞, 59.7)
11. 950, .8714
b. (.752, 1.100)
39. a. Yes
b. (6.45, 98.01)
c. (18.63, 85.83)
41. All 70%; (c), because it is shortest
43. a. 18.307
b. 3.940
c. .95
d. .10
45. (3.6, 8.1); no
13. a. (608.58, 699.74)
b. 189
47. a. 95% CI: (6.702, 9.456)
15. a. 80%
c. 75%
49. a. There appears to be a slight positive skew in the middle
half of the sample, but the lower whisker is much longer
than the upper whisker. The extent of variability is rather
substantial, although there are no outliers.
b. Yes. The pattern of points in a normal probability plot is
reasonably linear.
c. (33.53, 43.79)
b. 98%
17. 134.53
19. (.513, .615)
21. .218
23. a. (.438, .814)
25. a. 381
29. a. 2.228
e. 2.485
b. 659
b. 339
b. 2.086
f. 2.571
51. a. (.624, .732)
c. 2.845
d. 2.680
31. a. 1.812
b. 1.753
c. 2.602
d. 3.747
e. 2.1716 (from MINITAB)
f. Roughly 2.43
33. a. Reasonable amount of symmetry, no outliers
b. Yes (based on a normal probability plot)
c. (430.5, 446.1), yes, no
b. 1080
b. (.166, .410)
c. No
53. (Ϫ.84, Ϫ.16)
55. 246
57. (2tr /21Ϫ␣/2,2r, 2tr /2␣/2,2r) ϭ (65.3, 232.5)
59. a. (max(xi)/(1 Ϫ ␣/2)1/n, max(xi)/(␣/2)1/n)
c. (b); (4.2, 7.65)
b. (max(xi), max(xi)/␣1/n)
61. (73.6, 78.8) versus (75.1, 79.6)
Chapter 8
1. a. Yes
d. Yes
b. No
e. No
c. No
f. Yes
5. H0: ϭ .05 versus Ha: Ͻ .05. I: conclude variability in
thickness is satisfactory when it isn’t. II: conclude variability in thickness isn’t satisfactory when in fact it is.
7. I: concluding that the plant isn’t in compliance when it is; II:
concluding that the plant is in compliance when it isn’t.
9. a. R1
b. I: judging that one of the two companies is
favored over the other when that is not the case; II: judging
that neither company is favored over the other when in fact
one of the two really is preferred.
c. .044
d. (.3) ϭ (.7) ϭ .488, (.4) ϭ (.6) ϭ .845
e. Reject H0 in favor of Ha.
b. .01
11. a. H0: ϭ 10 versus Ha: 10
c. .5319, .0078
d. 2.58
e. 10.1032 is replaced by 10.124, and 9.8968 is replaced by
9.876.
f. xෆ ϭ 10.020, so H0 should not be rejected.
g. z Ն 2.58 or Յ Ϫ2.58
13. b. .0004, 0, less than .01
15. a. .0301
b. .003
c. .004
17. a. z ϭ 2.56 Ն 2.33, so reject H0.
d. .0052
19. a. z ϭ Ϫ2.27, so don’t reject H0.
21. a.
b.
c.
d.
b. .8413
b. .2266
c. 143
c. 22
t.025,12 ϭ 2.179 Ͼ 1.6, so don’t reject H0: ϭ .5.
Ϫ1.6 Ͼ Ϫ2.179, so don’t reject H0.
Don’t reject H0.
Reject H0 in favor of Ha: .5.
23. t ϭ 2.24 Ն 1.708, so H0 should be rejected. The data does
suggest a contradiction of prior belief.
25. a. z ϭ Ϫ3.33 Յ Ϫ2.58, so reject H0.
b. .1056
c. 217
x ϭ .640, s ϭ .3025, fs ϭ .480. A boxplot
27. a. ෆx ϭ .750, ~
shows substantial positive skew; there are no outliers.
b. No. A normal probability plot shows substantial curvature. No, since n is large.
700
Answers to Selected Odd-Numbered Exercises
c. z ϭ Ϫ5.79; reject H0 at any reasonable significance
level; yes.
d. .821
29. a. .498 Ͻ 1.895, so do not reject H0.
b. .72
31. Ϫ1.24 Ͼ Ϫ1.397, so prior belief does not appear to be
contradicted.
57. t Ϸ 1.9, so P-value Ϸ .041. Since P-value Յ ␣, H0: ϭ 25
should be rejected in favor of Ha: Ͼ 25.
59. t Ϸ 1.9, so P-value Ϸ .116. H0 should therefore not be
rejected.
b. P-value Ϸ 0. Yes.
61. a. .8980, .1049, .0014
35. Yes, because Ϫ2.47 Յ Ϫ1.96.
63. z ϭ Ϫ3.12 Յ Ϫ1.96, so H0 should be rejected.
37. z ϭ 3.67 Ն 2.58, so reject H0: p ϭ .40. No.
65. a. H0: ϭ .85 versus Ha: .85
b. H0 cannot be rejected for either ␣.
c. No
39. a. H0: p ϭ .02 vs Ha : p Ͻ .02, z ϭϪ1.01 Ͼ Ϫ1.645, don’t
reject H0, carry out inventory.
b. .1949
c. Ϸ 0
67. a. Yes, because t ϭ 12.9 Ն 2.228.
b. Normal population distribution
41. a. z ϭ 3.07 Ն 2.58, reject H0 and the company’s premise.
b. .0332
69. a. No; no
b. No, because z ϭ .44 and P-value ϭ .33 Ͼ .10.
43. No, no, yes. R ϭ {5, 6, . . . , 24, 25}, ␣ ϭ .098,  ϭ .090
71. a. Approximately .6; approximately .2 (from Appendix
Table A.17)
b. n ϭ 28
45. a. Reject H0.
c. Don’t reject H0.
e. Don’t reject H0.
47. a. .0778
d. .0066
b. Reject H0.
d. Reject H0. (a close call)
b. .1841
e. .5438
49. a. 0.40
e. Ͻ.005
73. a. z ϭ 1.64 Ͻ 1.96, so H0 cannot be rejected; Type II
b. .10. Yes.
75. Yes. z ϭ Ϫ3.32 Յ Ϫ3.08, so H0 should be rejected.
c. .0250
77. No, since z ϭ 1.33 Ͻ 2.05.
b. .018
c. .130
f. Ϸ .000
d. .653
79. P-value Ϸ 0, so reject H0; it appears that Ͼ 15.
51. P-value Ͼ ␣, so don’t reject H0; no apparent difference.
81. a. .01 Ͻ P-value Ͻ .025, so do not reject H0; no extradiction
53. P-value Ͻ .0004 Ͻ .01, so H0: ϭ 5 should be rejected in
favor of Ha: 5.
83. a. For H2: Ͻ 0, reject H0 if z Α xi /0 Յ 21Ϫ␣, 2n
b. Test statistic value ϭ 19.65 Ͼ 8.260, so do not reject H0.
55. No; P-value Ϸ .2
85. a. Yes, ␣ ϭ .002
Chapter 9
1. a. Ϫ.4 hr; it doesn’t
b. .0724, .2691
c. No
3. z ϭ 1.76 Ͻ 2.33, so don’t reject H0.
5. a. z ϭ Ϫ2.90, so reject H0.
c. .8212
d. 66
b. .0019
7. Yes, since z ϭ 1.83 Ն 1.645.
9. a. 6.2; yes
b. z ϭ 1.14, P-value Ϸ .25, no
c. No
d. A 95% CI is (10.0, 21.8).
29. t ϭ Ϫ2.10, df ϭ 25, P-value ϭ .023. At significance level
.05, we would conclude that cola results in a higher average
strength, but not at significance level .01.
31. a. Virtually identical centers, substantially more variability in
medium range observations than in higher range observations
b. (Ϫ7.9, 9.6), based on 23 df; no
33. t ϭ 1.33, P-value ϭ .094, don’t reject H0, no
11. A 95% CI is (.99, 2.41).
35. t ϭ Ϫ2.2, df ϭ 16, P-value ϭ .021 Ͼ .01 ϭ ␣, so don’t
reject H0.
13. 50
37. a. (Ϫ.561, Ϫ.287)
15. b. It increases.
17. a. 17
b. 21
c. 18
d. 26
19. t ϭ Ϫ1.20 Ͼ Ϫt.01,9 ϭ Ϫ2.821, so do not reject H0.
21. Yes; Ϫ2.64 Յ Ϫ2.602, so reject H0.
23. b. No
c. t ϭ Ϫ.38 Ͼ Ϫt␣/2,10 for any reasonable ␣, so
don’t reject H0 (P-value Ϸ .7).
25. (.3, 6.1), yes, yes
27. (6.5, 31.3) based on 9 df; yes, yes
b. Between Ϫ1.224 and .376
39. a. Yes
b. t ϭ 2.7, P-value ϭ .018 Ͻ .05 ϭ ␣, so H0 should be
rejected.
41. t ϭ 1.9, P-value ϭ .047. H0 cannot be rejected at significance level .01, but is barely rejected at ␣ ϭ .05.
43. a. No
b. Ϫ49.1
c. 49.1
45. a. 95% CI: (Ϫ2.52, 1.05); plausible that they are identical
b. Linear pattern in npp implies normality of difference distribution is plausible.
Answers to Selected Odd-Numbered Exercises
47. H0 is rejected because Ϫ4.18 Յ Ϫ2.33
69. (Ϫ299.3, 1517.9)
49. P-value ϭ .4247, so H0 cannot be rejected.
71. (1024.0, 1336.0), yes
51. a. z ϭ .80 Ͻ 1.96, so don’t reject H0.
b. n ϭ 1211
53. a. The CI for ln() is ln(ˆ) Ϯ z␣/2[(m Ϫ x)/(mx) ϩ
(n Ϫ y)/(ny)]1/2. Taking the antilogs of the lower and upper
limits gives a CI for itself.
b. (1.43, 2.31); aspirin appears to be beneficial.
73. Yes. t ϭ Ϫ2.25, df ϭ 57, P-value Ϸ .028
55. (Ϫ.35, .07)
57. a. 3.69
e. 4.30
b. 4.82
f. .212
c. .207
g. .95
d. .271
h. .94
701
75. a. No. t ϭ Ϫ2.84, df ϭ 18, P-value Ϸ .012
b. No. t ϭ Ϫ.56, P-value Ϸ .29
77. Not at significance level .05. t ϭ Ϫ1.76 Ͼ Ϫt.05,4 ϭ Ϫ2.015
79. No, nor should the two-sample t test be used, because a normal probability plot suggests that the good-visibility distribution is not normal.
81. Unpooled: df ϭ 15, t ϭ Ϫ1.8, P-value Ϸ .092
Pooled: df ϭ 24, t ϭ Ϫ1.9, P-value Ϸ .070
59. f ϭ .384; since .167 Ͻ .384 Ͻ 3.63, don’t reject H0.
61. f ϭ 2.85 Ն 2.08, so reject H0; there does appear to be more
variability in low-dose weight gain.
63. (s22 F1Ϫ␣/2 /s12 , s22 F␣/2 /s12 ); (.023, 1.99)
65. No. t ϭ 3.2, df ϭ 15, P-value ϭ .006, so reject H0: 1 Ϫ 2 ϭ
0 using either ␣ ϭ .05 or .01.
67. z Ͼ 0 d P-value Ͼ .5, so H0: p1 Ϫ p2 ϭ 0 cannot be
rejected.
83. a. m ϭ 141, n ϭ 47
b. m ϭ 240, n ϭ 160
85. z ϭ .83, P-value Ϸ .20, no.
87. .9015, .8264, .0294, .0000; true average IQs; no
89. Yes; z ϭ 4.2, P-value Ϸ 0
91. a. Yes. t ϭ Ϫ6.4, df ϭ 57, and P-value Ϸ 0
b. t ϭ 1.1, P-value ϭ .14, so don’t reject H0.
93. (Ϫ1.29, Ϫ.59)
Chapter 10
1. a. f ϭ 1.85 Ͻ 3.06 ϭ F.05,4,15, so don’t reject H0.
b. P-value Ͼ .10
19. Any value of SSE between 422.16 and 431.88 will work.
3. f ϭ 1.30 Ͻ 2.57 ϭ F.10,2,21, so P-value Ͼ .10. H0 cannot be
rejected at any reasonable significance level.
5. f ϭ 1.73 Ͻ 5.49 ϭ F.01,2,27, so the three grades don’t appear
to differ.
7. f ϭ 1.70 Ͻ 2.46 ϭ F.10,3,16, so P-value Ͼ .10. H0 cannot be
rejected at any reasonable significance level.
9. f ϭ 3.96 and F.05,3,20 ϭ 3.10 Ͻ 3.96 Ͻ 4.94 ϭ F.01,3,20, so
.01 Ͻ P-value Ͻ .05. Thus H0 can be rejected at significance
level .05; there appear to be differences among the grains.
11. w ϭ 36.09
3
1
4
2
5
437.5
462.0
469.3 512.8 532.1
Brands 2 and 5 don’t appear to differ, nor does there appear
to be any difference between brands 1, 3, and 4, but each
brand in the first group appears to differ significantly from
all brands in the second group.
13.
21. a. f ϭ 22.6 and F.01,5,78 Ϸ 3.3, so reject H0.
b. (Ϫ99.16, Ϫ35.64), (29.34, 94.16)
23.
1
2
3
4
1
Ϫ
Ϫ
Ϫ
Ϫ
2
3
2.88 Ϯ 5.81 7.43 Ϯ 5.81
Ϫ
4.55 Ϯ 6.13
Ϫ
Ϫ
Ϫ
Ϫ
4
3
2
4
12.78 Ϯ 5.48
9.90 Ϯ 5.81
5.35 Ϯ 5.81
Ϫ
1
25. a. Normal, equal variances
b. SSTr ϭ 8.33, SSE ϭ 77.79, f ϭ 1.7, H0 should not be
rejected (P-value Ͼ .10)
27. a. f ϭ 3.75 Ն 3.10 ϭ F.05,3,20, so brands appear to differ.
b. Normality is quite plausible (a normal probability plot of
the residuals xij Ϫ ෆxiи shows a linear pattern).
c. 4 3 2 1 Only brands 1 and 4 appear to differ
significantly.
3
1
4
2
5
427.5
462.0
469.3
502.8
532.1
ෆnෆ)
33. arcsin(͙x/
4
33.84
35. a. 3.68 Ͻ 4.94, so H0 is not rejected.
b. .029 Ͼ .01, so again H0 is not rejected.
15. w ϭ 5.94
2
24.69
1
26.08
3
29.95
The only significant differences are between 4 and both
1 and 2.
17. (Ϫ.029, .379)
31. Approximately .62
37. f ϭ 8.44 Ͼ 6.49 ϭ F.001, so P-value Ͻ .001 and H0 should
be rejected.
5 3 1 4 2 This underscoring pattern is a bit awkward
to interpret.