PROBABILITY AND STATISTICS

Homework – Chapter 8: Tests of Hypotheses Based

on one or two Samples (Cont.)
Sec Ex Solution Comment
. .
8.2 I 15 a. P(z 1.88 when z has a standard normal distribution)
(8) 11.880.0301
b. P(z 2.75 when z ~ N(0, 1) 2.750.003
c. 2.8812.880.004
16 a. P(t 3.733 when t has a t distribution with 15 d.f.) = 0.001, because
the 15 d.f. row of Table A.5 shows that t.001,15 = 0.3733
b. d.f. = n – 1 = 23, so P(t 2.500) 0.01
c. d.f. = 30, and P(t 1.697) P(t 1.697) = 0.05 + 0.05 = 0.10
II
17-19 18 a. (72.3-75)/1.8= - 1.5 so 72.3 is 1.5 SD’s (of x́ ) below 75.
al. 2 b. Ho is rejected if z 2.33; since z 1.5 is not 2.33, don’t reject Ho.
c. area under standard normal curve below –2.88 2.880.0020
d. 2.88 + (75-70)/(9/5)) = 0.1) = 0.4602 so 700.5398
e. n = [ 9(2.88+2.33)2/(75-70)]2 = 87.95 , so use n = 88
f. 76P(Z 2.33 when 76) P(X 72.9 when 76) = 72.9 –
76)/0.9) = 3.44) = 0.0003
19 a. Reject Ho if either z 2.58 or z 2.58;  /√ n =0.3 , so
z = (94.32-95)/0.3 = -2.27 . Since –2.27 is not < -2.58, don’t reject Ho.
b.
94

c. n = [(1.20(2.58+1.28) )/(95 – 94)]2 = 21.46, so use n = 22.
20-25 20 With Ho: 750 , and Ha: 750 and a significance level of .05, we
al. 4 reject Ho if z < -1.645; z = -2.14 < -1.645, so we reject the null hypothesis and
do not continue with the purchase. At a significance level of .01, we reject Ho
if z < -2.33; z = -2.14 > -2.33, so we don’t reject the null hypothesis and thus
continue with the purchase.
21 With Ho: 0.5 , and Ha: 0.5 we reject Ho if t t/ 2,n1 or t t/
2,n1
a. 1.6 < t.025,12 = 2.179, so don’t reject Ho
b. -1.6 > -t.025,12 = -2.179, so don’t reject Ho
c. – 2.6 > -t.005,24 = -2.797, so don’t reject Ho
d. –3.9 < the negative of all t values in the df = 24 row, so we reject Ho in
favor of Ha.
22 a. It appears that the true average weight could be more than the production
specification of 200 lb per pipe.
b. Ho: 200 , and Ha: 200 we reject Ho if t t.05,29 1.699 .
t = (206.73 -200)/ (6.35/√ 30 ) = 5.80 > 1.699, so reject Ho.
The test appears to substantiate the statement in part a.
23 Ho: 360 vs. Ha: 360 ; t =( x́ – 360)/(s/√ n )
reject Ho if t t.05,25 1.708 ;
t = ((370.69 – 360)/ (24.36/√ 26 )) = 2.24 > 1.708.
Thus Ho should be rejected. There appears to be a contradiction of the prior
 belief.
8.3 I
(8) 35-38 35 1 Parameter of interest: p = true proportion of cars in this particular county
passing
al. 3 emissions testing on the first try.
2 Ho: p = 0.70
3 Ha: p 0.70
4 z = (( ^p – p0)/ √ p 0(1− p 0)/n ) = (( ^p – 0.70)/ √ 0.70(0.30)/n )
5 either z 1.96 or z -1.96
6 z = (( 124/200– 0.70)/ √ 0.70(0.30)/200 ) = -2.469
7 Reject Ho. The data indicates that the proportion of cars passing the first
time on emission testing or this county differs from the proportion of cars
passing statewide
37 1 p = true proportion of the population with type A blood
2 Ho: p = 0.40
3 Ha: p 0.40
4 z = (( ^p – p0)/ √ p 0(1− p 0)/n ) = (( ^p – 0.40)/ √ 0.40(0.60)/n )
5 Reject Ho if z 2.58 or z -2.58
6 z = (( 82/150– 0.40)/ √ 0.40(0.60)/150 ) = 3.667
7 Reject Ho. The data does suggest that the percentage of the population
with type A blood differs from 40%. (at the 0.01 significance level). Since the
z critical value for a significance level of .05 is less than that of .01, the
conclusion would not change.
38 a. Let the number of couples tended to learn more to the right than to the left
be x = 80
The propotion of right-learning behavior be p = 2/3
Let the number of couples in a sample be n = 124
The expected values of X is,
E(X) = np
= 1248(2/7)
= 82.666667
A random sample of 124 kissing couples, both people in 80 of the couple
tended to learn more to the right than to the left
Here, the researcher observe that x= 80 for a difference of 2.666667
So,
P(| X−82.666667|2.666667) = P(X80 or X85.33)
= 0.634152
b. Test the claim that proportion of right-learning behavior differ from 2/3)
The null and alternative hypotheses are,
Ho: p = 2/3
Ha: p 2/3
z = (( ^p – p0)/ √ p 0(1− p 0)/n ) = ((80/124– 2/3 √ 2/3(1/3)/124 ) =
-0.51
There are not enough evidence to conclude that the rue propotion of right-
learning behavior differs from 2/3
II 39 a. We wish to test Ho: p = .02 vs Ha: p < .02; only if Ho can be rejected will
the inventory be postponed. The lower-tailed test rejects Ho if z -1.645
With ^p = 15/1000 – 0.015 , z = -1.01, which is not -1.645. Thus, Ho
cannot be rejected, so the inventory should be carried out.
b. 0.01) = 0.02 -0.01 + 1.645 √ 0.02(0.98)/1000 )/ (
√ 0.01(0.99)/1000 ]= = 1
 c. 0.05) = 0.02 -0.05 + 1.645 √ 0.02(0.98)/1000 )/ (
√ 0.05(0.95)/1000 ]= = 0.0005, so is p = 0.05 it is highly
unlikely that Ho will be rejected and the inventory will almost surely be
carried out.
40 Let p denote the true proportion of those called to appear for service who are
black. We wish to test Ho: p = 0.25 vs Ha: p < 0.25
We use z = (( ^p – 0.25)/ √ 0.25(0.75)/n ) with the rejection region z -
z.01 = -2.33. We calculate ^p = 177/1050 = 0.1686, and z = (0.1686 -
0.25)/0.0134 = -6.1. Because –6.1 < -2.33, Ho is rejected. A conclusion that
discrimination exists is very compelling.
41 a. P = true proportion of current customers who qualify. Ho: p = 0.05 vs Ha:
p 0.05
z = (( ^p – 0.05)/ √ 0.05(0.95)/n ) , reject Ho if z 2.58 or z -2.58. pˆ
0.08, so z = 0.03/0.00975 – 3.07 so Ho is rejected. The company’s
premise is not correct
b. 0.10) = 0.05 -0.10 + 2.58 √ 0.05(0.95)/500 )/ (
√ 0.10(0.90)/500 ] = 
42 a. The alternative of interest here is Ha: p > 0.50 (which states that more than
50% of all enthusiasts prefer gut), so the rejection region should consist of
large values of X (an upper-tailed test). Thus (15, 16, 17, 18, 19, 20) is the
appropriate region.
b. P(15 X when p 0.5) = 1 – B(14; 20, 0.05) = 0.021, so this is a level
0.05 test. For R = {14, 15, …, 20}, = 0.058, so this R does not specify a
level 0.05 test and the region of a is the best level 0.05 test. (0.05 along
with smallest possible ).
c. (0.6) = B(14; 20, 0.6) = 0.874, and (0.8) = B(14; 20, 0.8) = 0.196.
d. The best level 0.10 test is specified by R = (14, …, 20} (with = 0.052)
Since 13 is not in R, Ho is not rejected at this level.
43 The hypotheses are Ho: p = 0.10 vs. Ha: p > 0.10, so R has the form {c, …,
n}. For n = 10, c = 3 (i.e. R = {3, 4, …, 10}) yields = 1 – B(2; 10, 0.1) =
0.07 while no larger R has 0.10; however (0.3) = B(2; 10, 0.3) = 0.383.
For n = 20, c = 5 yields = 1 – B(4; 20, 0.1) = 0.043, but again (0.3) = B(4;
20, 0.3) = 0.238. For n = 25, c = 5 yields = 1 – B(4; 25, 0.1) = 0.098 while
(0.7) = B(4; 25, 0.3) = 0.090 .10, so n = 25 should be used.
8.4 I 45 Using = 0.05, Ho should be rejected whenever p-value < 0.05.
a. P-value = 0.001 < 0.05, so reject Ho
(5) b. .021 < 0.05, so reject Ho.
c. 0.078 is not < 0.05, so don’t reject Ho.
d. .047 < 0.05, so reject Ho ( a close call).
e. .148 > 0.05, so Ho can’t be rejected at level .05.
46 a. p-value = 0.084 > 0.05 = , so don’t reject Ho.
b. p-value = 0.003 < 0.001 = , so reject Ho.
c. 0.498 >> 0.05, so Ho can’t be rejected at level 0.05
d. 0.084 < 0.10, so reject Ho at level 0.10
e. 0.039 is not < 0.01, so don’t reject Ho.
f. p-value = 0.218 > 0.10, so Ho cannot be rejected.
47 a. z = 1.42
p – value = 1 - z)
= 1 - 1.42)
= 0.0778
b. z = 0.9
p – value = 1 - z)
 = 1 - 0.9)
= 0.1814
c. z = 1.96
p – value = 1 - z)
= 1 - 1.96)
= 0.025
d. z = 2.48
p – value = 1 - z)
= 1 - 2.48)
= 0.0066
e. z = -0.11
p – value = 1 - z)
= 1 - -0.11)
= 0.5438
48 a. z = 2.10
p – value = 2(1 - z))
= 2(1 - 2.10))
= 0.0357
b. z = -1.75
p – value = 2(1 - z))
= 2(1 - 1.75))
= 0.0801
c. z = -0.55
p – value = 2(1 - z))
= 2(1 - 0.55))
= 0.5823
d. z = 1.41
p – value = 2(1 - z))
= 2(1 - 1.41))
= 0.1585
e. z = -5.3
p – value = 2(1 - z))
= 2(1 - 5.3))
= 0.0000
II
49-50 49 a. In the df = 8 row of table A.5, t = 2.0 is between 1.860 and 2.306, so the p-
value is between 0.025 and 0.05: 0.025 < p-value < 0.05.
al. 1 b. 2.201 < | -2.4 | < 2.718, so 0.01 < p-value < 0.025.
c. 1.341 < | -1.6 | < 1.753, so 0.05 < P( t < -1.6) < 0.10. Thus a two-tailed p-
value: 2(0.05 < P( t < -1.6) < 0.10), or 0.10 < p-value < 0.20
d. With an upper-tailed test and t = -0.4, the p-value = P( t > -0.4) >0.50.
e. 4.032 < t=5 < 5.893, so 0.001 < p-value < 0.005
f. 3.551 < | -4.8 |, so P(t < -4.8) < 0.0005. A two-tailed p-value = 2[ P(t <
-4.8)] < 2(0.0005), or p-value < 0.001.