Vous êtes sur la page 1sur 15

Basic and Applied Questions- HYPOTHESIS TESTING

78. The Daytona Beach Tourism Commission recently claimed that the average amount of
money a typical college student spends per day during spring break is over $70. Based
upon previous research, the population standard deviation is estimated to be $17.32.
The Commission surveys 35 students and find that the mean spending is $67.57. Is there
evidence that the average amount spent by students is less than $70?

ANSWER:
H 0 :   70 vs. H 1 :  < 70
Test statistic Z = (67.57 – 70.00) / (17.32 / 5.92) = -0.83.
Z= xbar – u / standard deviation/ sqrt(n).
Since p-value = P(Z < -0.83) = 0.2033, we fail to reject H 0 at all  levels less than
20.33%. There is no sufficient evidence to disprove the claim that the average amount
spent by students is over $70.

79. The supervisor of a production line believes that the average time to assemble an
electronic component is 14 minutes. Assume that assembly time is normally distributed
with a standard deviation of 3.4 minutes. The supervisor times the assembly of 14
components, and finds that the average time for completion was 11.6 minutes. Is there
evidence that the average amount of time required to assemble a component is
something other than 14 minutes? Use  = 0.01

ANSWER:
H 0 :  =14 vs. H 1 :   14
Test statistic Z= (11.6 - 14) / (3.4 / 6.32) = -4.46
Sample mean = 11,6 ; population mean = 14 ; standard deviation 3.4 ; n = 14
Since p-value = 2P(Z < -4.46)  0.0041, we reject H 0 at  = 0.01. We conclude that
there is sufficient evidence to disprove the claim that the average time to assemble an
electronic component is 14 minutes (The population mean appears to be less than 14
minutes.)
0.0041 is bigger than 0.01 that s why we reject H0.

80. The manufacturer of a new product claims that his product will increase output per
machine by at least 29 units per hour. A line manager adopts the product on 15 of his
machines, and finds that the average increase was only 26 with a standard deviation of
6.2. Is there evidence to doubt the manufacturer’s claim?

ANSWER:
H 0 :   29 vs. H 1 :  < 29
Test statistics t = (26 - 29) / (6.2 / 1.6) = -0.77
The standard deviation is unkown so we apply t test.
T = xbar – u0 / s / sqrt(n)
Reject H 0 if t < - tn 1,   t14,0.10 = -1.345.
Since t = -0.77, we are unable to reject H 0 at  = 0.10. There is insufficient evidence to
disprove claim that the product will increase output per machine by 29 units per hour.

81. The manufacturer of a certain chewing gum claims that four out of five dentists surveyed
prefer their type of gum. You decide to test their claim. You find that in a sample of 200
doctors, 74.1% do actually prefer their gum. Is this evidence sufficient to doubt the
manufacturer’s claim? Use  = 0.025
ANSWER:

263
Hypothesis Testing
H 0 : P = 0.80 vs. H 1 : P < 0.80
Test statistics Z =  pˆ  P0  / P0 (1  P0 ) / n  2.09
Since p-value = P(Z < -2.09) = 0.0183, we reject H 0 at  = 0.025. There is sufficient
evidence to doubt the manufacturer’s claim.
0.0183 is smaller than 0.025 th ats why we rejects

82. The U.S. Postal Service advertises that 63.4% of all third class mail is read by the
recipients. An environmental group, concerned about waste generation, wants to test this
assertion. They take a sample of 220 households, and find that only 58.7% read their
third class mail. Should you believe the U.S. Postal Service claim at the 10% level of
significance?

ANSWER:
H 0 : P = 0.634 vs. H 1 : P < 0.634
Test statistic Z =  pˆ  P0  / P0 (1  P0 ) / n  1.38
X= 0.587 * 220= 129.14
p = 129.14 / 220 = 0.587
Z = 0.587 / sqrt (0.634 * 0.466 / 220) = -1.38
Since p-value = P(Z < -1.38) = 0.0838, we reject H 0 at  = 0.10. There is sufficient
evidence to doubt the U.S. Postal Service claim
0.0838 is smaller than 0.10

83. The manufacturer of bags of cement claims that they fill each bag with at least 50.2
pounds of cement. Assume that the standard deviation for the amount in each bag is 1.2
pounds. The decision rule is adopted to shut down the filling machine if the sample mean
weight for a sample of 40 bags is below 49.8. Suppose that the true mean weight of
bags is 50 pounds. Using this decision rule, what is the probability of a Type II error?

ANSWER:
  P  X  X crit |       P  X  49.8 |   50  = P(Z > -1.05) = 0.8531

84. An accountant claims that he can complete a standard tax return in at most an hour. For
a random sample of 24 tax returns, the accountant averaged 63.2 minutes with a
standard deviation of 7.7 minutes. Is there sufficient evidence to suggest that the
accountant’s claim is incorrect?

ANSWER:
H 0 :   60 vs. H 1 :  > 60
Test statistic t = (63.2 - 60) / (7.7/4.9) = 2.04
Standard deviation is unknown
t = xbar – u / s / sqrt(n) = 2.04
Since, t23,0.05  1.714 , we reject H 0 at   0.05. There is sufficient evidence to disprove
the accountant claim that he can complete a standard tax return in at most an hour.

85. As shift manager at a local fast food place, you are responsible for ensuring quality
control. You do not want to weigh all the frozen hamburger patties that get delivered by
your supplier to make sure they weigh four ounces on average, so you have one of your
minimum wage earners do it. Assume that the standard deviation of the weight of
hamburger patties is known to be 0.1 ounces. You tell your employee to choose 25
patties at random and find the average weight. For what values would you tell the
employee to reject the shipment, if you wanted the probability of a Type I error to be 0.10
or less?

264
Hypothesis Testing

ANSWER:
H 0 :  = 4.0 vs. H 1 :  < 4.0
Reject H 0 if Z = ( X – 4.0) / (0.1/5) < - Z 0.10 = -1.28; that is, reject shipments for X < 4.0
-1.28(0.02) = 3.9744

86. An assembly line will be shut down for maintenance if the defect rate exceeds 2.3%.
Suppose you adopt a 5% significance level and take a random sample 200 items off the
assembly line and compute the proportions that are defective. For what values of the
sample proportion will the assembly line be shut down?

ANSWER:
H 0 : P = 0.023 vs. H 1 : P > 0.023
Reject H 0 if  pˆ  P0  /  P0 1  P0  / n  > Z.05 = 1.645, that is, assembly line will be shut
down for maintenance if ( p̂ -0.023) / 0.0106 > 1.645, or equivalently if sample proportion
exceeds 0.0404

87. The state lottery office claims that the average household income of those people playing
the lottery is greater than $37,000. Assume that the distribution of household income of
those people playing the lottery is normally distributed with a standard deviation of
$5,756. Suppose that for a sample of 25 households, it is found that the average income
was $36,243. Is the state lottery incorrect in asserting that the average household income
of lottery players exceeds $37,000? Use the 10% significance level.

ANSWER:
H 0 :   37,000 vs. H 1 :  < 37,000
Test statistic Z = (36,243 – 37,000) / (5,756 / 5) = -0.658
Z= xbar – u / standard deviation/ sqrt(n).
Since, p-value = P(Z < -0.66) = 0.2546, we fail to reject H 0 at  = 0.10. There is no
sufficient evidence to disprove the claim of the state lottery office.
0.2546 is greater than the significance level .

88. A professor claims that the average score on a recent exam was 83. Assume that the
test scores are normally distributed. You believe that the average score is less than 83,
therefore, you ask some people in class how they did, and you record the following
scores: 82, 77, 85, 76, 81, 91, 70, and 82. Do you have any reason at the 10% level of
significance to think the professor’s statement was incorrect?

ANSWER:
H 0 :  = 83 vs. H 1 :  < 83
The sample mean and standard deviation are x = 80.5 and s = 6.3, respectively.
Test statistic t = (80.5 - 83) / (6.3 / 2.83) = -1.12
Standard deviation is unknown we apply t test:
Since, tn 1,  / 2  t7,.10  1.415 , we fail to reject H 0 at  = 0.10. There is insufficient
evidence to disprove the claim.
We used the t table

265
Hypothesis Testing

89. A local transportation planning group is concerned about the lack of car-pooling on the
part of commuters. They are afraid that the proportion of local drivers car-pooling is below
the national average of 20%. A survey of 356 local drivers reveals that 18.7% of them car
pool. Is there evidence that the actual proportion of local commuters car-pooling is less
than 20%? Test at the 10% level of significance.

ANSWER:
H 0 : P = 0.20 vs. H 1 : P < 0.20
Test statistic Z =  pˆ  P0  / P0 (1  P0 ) / n = -0.62
Z = (0.187 – 0.20) / sqrt ( 0.20 * 0.80 / 356) = 0.62

Since p-value = P(Z < -0.62) = 0.2676, we fail to reject null hypothesis at  = 0.10. There
is insufficient evidence that the actual proportion of local commuters car-pooling is less
than 20%.

We ve used Z table and obtained 0.2676 wich is greater than 0.10

90. Suppose that you want to test H 0 :  = 277 vs. H1 :   277 at  = 0.05, and you know
that the population standard deviation  = 13.5. If you randomly select a sample of 20
observations, for what values of sample mean will you reject the null hypothesis?

ANSWER:
H 0 :  =277 vs. H 1 :   277
Reject H 0 if Z = (k - 277) / (13.5 / 4.472) < - Z 0.05 = -1.645 or Z > Z 0.05 = 1.645. That is,
reject H 0 for sample mean x  k < 277 -1.645(3.019) = 272.03 or sample mean x  k >
277 + 1.645(3.019) = 281.97.

91. A manufacturing manager is seeking strong evidence to support his hope that new
operation procedures have reduced the percentage of under-filled water bottles from the
Stanwood, Michigan, Ice Mountain Water Plant. State his null and alternative hypotheses
and indicate the results that would provide strong evidence.

ANSWER:
H 0 : Pnew  Pold (There is no difference in the percentage of under-filled water bottles
between new and old operating procedures)
H1 : Pnew  Pold (The percentage of under-filled water bottles has been reduced with the
new operating procedures)

QUESTIONS 92 THROUGH 94 ARE BASED ON THE FOLLOWING INFORMATION:


The battery pack of a hand calculator is supposed to perform 20,000 calculations before needing
recharging. The quality – control manager for the manufacturer is concerned that the pack may
not be working for as long as the specifications state. A test of 114 battery packs gives an
average of 19,695 calculations and a standard deviation of 1103.

92. Formulate the appropriate null and alternative hypotheses.

ANSWER:
H 0 :   20,000, H1 :   20,000.

93. Calculate the appropriate test statistic and p – value.

266
Hypothesis Testing
ANSWER:
t  ( x  0 ) /( s / n )  (19,695  20,000) /(1,103/ 114)  2.952
Since df = n -1 = 113, and t,0.005  2.576 , then p – value < 0.005.

94. Is the result in Question 93 statistically significant at the usual  levels; namely 0.01,
0.05, and 0.10? Would you call the result practically significant?

ANSWER:
Since the p–value is less than  for the usual  = 0.01, 0.05, and 0.10, the result is
statistically significant at the usual  values. The result is not what we would call
“practically significant”. If the battery needs recharging after, say, 19,785 calculations
instead of 20,000, we seriously doubt that customers will return calculators for this
(19,785 is about 98.9% of 20,000 – this is not serious mis-advertising on the part of the
manufacturer.

QUESTIONS 95 THROUGH 98 ARE BASED ON THE FOLLOWING INFORMATION:


A team of builders has surveyed buyers of their new homes for years. Consistently, only 41% of
the buyers have indicated that they were “quite satisfied” or “very satisfied” with the construction
quality of their homes. The builders have adopted a revised quality – inspection system to try to
improve customer satisfaction. They have surveyed 104 buyers since then; these buyers seem
representative, with no systematic changes from past purchasers. Of the 104 buyers, 52
indicated they were quite or very satisfied.

95. Formulate the null hypothesis that there has been no real change in customer satisfaction
from the past rate.

ANSWER:
H 0 : P  0.41
H0 null hypothesis and 0.41 the percent of buyers

96. Before taking such a survey, would you use a one-sided or two-sided alternative
hypothesis?

ANSWER:
The exercise indicates that the purpose of the revised inspection system is to increase
customer satisfaction. We want to use the sample to see if that purpose has been
achieved. Therefore, we would use a one – sided alternative hypothesis; namely,
H1 : P  0.41

97. Calculate the appropriate statistic for testing the null hypothesis.

ANSWER:
pˆ  P0 0.50  0.41
pˆ  52 /104  0.50 ; then test statistic Z    1.87
P0 1  P0  / n (0.41)(0.59) /104

98. Test the null hypothesis in Question 95 at   0.05.

ANSWER:
Reject H 0 if Z  z0.05  1.645 . Since Z = 1.87, H 0 is rejected at   0.05 . We conclude
there is sufficient evidence that adopting the revised quality – inspection system has
improved the proportion of customer satisfaction.

267
Hypothesis Testing

QUESTIONS 99 AND 100 ARE BASED ON THE FOLLOWING INFORMATION:


During 2000 and 2001 many people in Europe objected to purchasing food that was genetically
modified, produced by farmers in the United States. The U.S. farmers argued that there was no
scientific evidence to conclude that these products were not healthy. The Europeans argued that
there still might be a problem with these foods.

99. State the null and alternative hypotheses from the perspective of the Europeans.

ANSWER:
H 0 : Genetically modified food produced by farmers in the United States was not safe
H1 : Genetically modified food produced by farmers in the United States was safe
100. State the null and alternative hypotheses from the perspective of the U.S. farmers.

ANSWER:
H 0 : Genetically modified food produced by farmers in the United States was safe
H1 : Genetically modified food produced by farmers in the United States was not safe
101. Explain carefully the distinction between simple and composite hypotheses

ANSWER:
A simple hypothesis assumes a specific value for the population parameter that is being
tested. A composite hypothesis assumes a range of values for the population parameter.

102. Explain carefully the distinction between one-sided and two-sided alternatives

ANSWER:
One sided alternatives can be either a one-tailed upper (greater than) or a one-tailed
lower (less than) statement about the population parameter. Two sided alternatives are
made up of both greater than or less than statements and are written as (not equal to).

QUESTIONS 103 THROUGH 106 ARE BASED ON THE FOLLOWING INFORMATION:


A pharmaceutical manufacturer is concerned that the mean impurity concentration in pills should
not exceed 2%. It is known that from a particular production run impurity concentrations follow a
normal distribution with standard deviation 0.32%. A random sample of 64 pills from a production
run was checked, and the sample mean impurity concentration was found to be 2.05%.

103. Test at the 5% level the null hypothesis that the population mean impurity concentration
is 2% or less against the alternative that it is more than 2%.

ANSWER:
H 0 :   2 and H1 :   2 .
Reject H 0 if Z  Z.05  1.645
2.05  2
Z  1.25 , therefore, we fail to reject H 0 at the 5% level. We conclude that
0.32 / 64
the population mean impurity concentration is 2% or less.
Z = xbar –u / st.deviation / sqrt(n) = 1.25

104. Calculate the p-value for this test.

ANSWER:
p -value = P( Z > 1.25) = 0.50 – 0.3944 = 0.1056

268
Hypothesis Testing
105. Suppose that the alternative hypothesis in Question 103 had been two-sided rather than
one-sided. State, without doing the calculations, whether the p-value of the test would be
higher than, lower than, or the same as that found in Question 104. Explain your
reasoning.

ANSWER:
The p-value would be higher (namely; 0.2112) – since the p-value now corresponds to
the area in both the lower and upper tails of the distribution whereas before it was the
area in the upper tail only.

106. In the context of this problem, explain why a one-sided alternative hypothesis is more
appropriate than a two-sided alternative.

ANSWER:
A one-sided alternative is more appropriate since we are not interested in detecting
possible low levels of impurity, only high levels of impurity.

107. Explain carefully the distinction between Type I and Type II errors

ANSWER:
A Type I error is falsely rejecting the null hypothesis. To commit a Type I error, the truth
must be that the null hypothesis is really true and yet you conclude to reject the null and
accept the alternative. A Type II error is falsely not rejecting the null hypothesis when in
fact the null hypothesis is false. To make a Type II error, the null hypothesis must be
false (the alternative is true) and yet you conclude to not reject the null hypothesis.

108. Explain carefully the distinction between significance level and power

ANSWER:
Significance level is the chosen level of significance that established the probability of a
making a Type I error. This is represented by alpha. The power of the test is the ability of
the hypothesis test to identify correctly a false null hypothesis and reject it.

QUESTIONS 109 THROUGH 112 ARE BASED ON THE FOLLOWING INFORMATION:


An engineering research center claims that, through the use of a new computer control system,
automobiles should achieve, on average, at least an additional 5 miles per gallon of gas. A
random sample of 81 automobiles was used to evaluate this product. The sample mean increase
in miles per gallon achieved was 4.5, and the sample standard deviation was 2.9 miles per gallon.

109 State the appropriate null and alternative hypotheses.

ANSWER:
H 0 :   5 and H1 :   5 .

110. Test the hypothesis at the 5% level.

ANSWER:
Reject H 0 if Z   Z.05  1.645

269
Hypothesis Testing
4.5  5.0
Z  1.41 , therefore, we fail to reject H 0 at the 5% level. We conclude that
3.2 / 81
through the use of a new computer control system, automobiles should achieve, on
average, at least an additional 5 miles per gallon of gas.
Z = xbar – u / deviation / sqrt(n) = -1.41

111. Find the p-value of this test, and interpret your findings.

ANSWER:
p -value = P( Z < -1.41) = 0.50 – 0.4207 = 0.0793. This means that H 0 would be rejected
for any significance level above 7.93%
we ve used Z- table.

112. Explain how to use the p-value in Question 111 to test the hypothesis in Question 109 at
the 5% level.

ANSWER:
Since p -value = 0.0793 >  = 0.05, we fail to reject H 0 at the 5% level.
.

QUESTIONS 113 THROUGH 117 ARE BASED ON THE FOLLOWING INFORMATION:


A company selling new econometrics computer software advertises that firms using this software
obtain, on average during the first year, a yield of at least 10% on their initial investments. A
random sample of 10 of these franchises produced the following yields for the first year of
operation: 9.4, 11.1, 10.4, 10.5, 11.2, 8.2, 8.5, 4.0, 8.8, and 6.1. Assume that population yields
are normally distributed.

113. State the appropriate null and alternative hypotheses.

ANSWER:
H 0 :   10 and H1 :   10 .

114. Calculate the sample mean and sample standard deviation.

ANSWER:
x  8.82 and s  2.301
x bar = 9.4 + 11.1 + 10.4 + 10.5+ 11.2 + 8.2 + 8.5 + 4.0 +8.8 + 6.1 / 10 = 8.82
s = (9.4 – 8.82)^2 + (11.1 – 8.82)^2+ (10.4 – 8.82)^2+ (8.2 – 8.82)^2+ (8.5 – 8.82)^2+
(4.0 – 8.82)^2+ (8.8 – 8.82)^2+ (6.1 – 8.82)^2+ (10.5 – 8.82)^2+ (11.2– 8.82)^2/ 9 = 2.301

115. Test the company’s claim at the 5% significance level.

ANSWER:
Reject H 0 if t  t9,.05  1.833
x  0 8.82  10
t   1.622 , therefore, we fail to reject H 0 . We conclude that firms
s/ n 2.301/ 10
using this new software obtain, on average during the first year, a yield of at least 10% on
their initial investments.

270
Hypothesis Testing

116. Use statistical software to verify your answers to Questions 114 and 115, and calculate
the p-value.

ANSWER:

Results for one-sample analysis for Data

Summary measures
Sample size 10
Sample mean 8.820
Sample standard deviation 2.301

Test of mean>=10 versus one-tailed alternative


Hypothesized mean 10.000
Sample mean 8.820
Std error of mean 0.728
Degrees of freedom 9
t-test statistic -1.622
p-value 0.070

117. Interpret the p-value reported in Question 116.

ANSWER:
Since p -value = 0.07, this means that H 0 would be rejected for any significance level
above 7%

QUESTIONS 118 AND 119 ARE BASED ON THE FOLLOWING INFORMATION:


On the basis of a random sample the null hypothesis: H 0 :   0 is tested against the
alternative: H1 :   0 and the null hypothesis is not rejected at the 10% significance level.

118. Does this necessarily imply that  0 is contained in the 90% confidence interval for  ?

ANSWER:
No, the 90% confidence level provides for 5% of the area in either tail. This does not
correspond to a one-tailed hypothesis test with  = 10% which has 10% of the area in
one of the tails.

119. Does this necessarily imply that  0 is contained in the 80% confidence interval for  if
the observed sample mean is larger than  0 ?

ANSWER:
Yes, the 80% confidence level provides for 10% of the area in either tail. This
corresponds to a one-tailed hypothesis test with  = 10% which has 10% of the area in
one of the tails.

QUESTIONS 120 AND 121 ARE BASED ON THE FOLLOWING INFORMATION:


In a random sample of 200 education school students, 92 sample members indicated some
measure of agreement with this statement: “Scores on a standardized entrance exam are less
important for a student’s chance to succeed academically than is the student’s high school GPA”.

271
Hypothesis Testing
120. Test at the 5% level of significance the null hypothesis that one-half of all education
school graduates would agree with this statement against a two-sided alternative.

ANSWER:
H 0 : P  0.5 and. H1 : P  0.5
p̂ = 92 / 200 = 0.46. Reject H 0 if Z   Z 0.025  1.96 or Z  Z 0.025  1.96
pˆ  P0 .46  .50
Z   1.13 , therefore, reject H 0 . We conclude
P0 (1  P0 ) / n (0.5)(0.5) / 200
that the proportion of education school graduates who would agree with the above
statement if different from 0.50.
We ve used the z-table

121. Find and interpret the p-value of the test.

ANSWER:
p -value = 2  P( Z < -1.13) = 2(0.50 – 0.3708) = 0.2584
The probability of finding a random sample with a sample proportion p̂ this far or further
from 0.5 if the null hypothesis is really true is 0.2584
We ve used the z-table

122. An automobile company is attempting to determine if it should retain a previously popular


sports car model. A random sample of 425 men is obtained, and each person in the
sample is asked if he would purchase this existing car model. To determine if the old car
model should be retained, the sample proportion of men who said yes, p̂ , is used at a
level  = 0.025 to test the hypothesis H 0 : P  0.25 vs. H1 : P  0.25 . What value of the
sample proportion, p̂ , is required to reject the null hypothesis?

ANSWER:
Reject H 0 if Z   Z.025  1.96
pˆ  P0 pˆ  0.25 pˆ  0.25 pˆ  0.25
Z   . Then =-1.96  pˆ  0.2088
P0 (1  P0 ) / n (0.25)(0.75) / 425 0.021 0.021

123. Carefully explain what is meant by the p-value of a test, and discuss the use of this
concept in hypothesis testing.

ANSWER:
The p-value indicates the likelihood of getting the sample result at least as far away from
the hypothesized value as the one that was found, assuming that the distribution is really
centered on the null hypothesis. The smaller the p-value, the stronger the evidence
against the null hypothesis

124. A random sample of 250 business faculty members was asked if there should be a
required foreign language course for international business majors. Of these sample
members, 170 felt there was a need for a foreign language course. Test the hypothesis
that at least 75% of all business faculty members hold this view. Use  = 0.05.

ANSWER:

272
Hypothesis Testing
H 0 : P  0.75 and. H1 : P  0.75
p̂ = 170 / 250 = 0.68. Reject H 0 if Z   Z 0.05  1.645
We ve used z table and find the value – 1.645

pˆ  P0 0.68  0.75
Z   2.56 , therefore, reject H 0 . We conclude
P0 (1  P0 ) / n (0.75)(0.25) / 250
that less than 75% of all business faculty members hold the view that there should be a
required foreign language course for international business majors.
-2.56 is greater than – 1.645 so we reject H0

QUESTIONS 125 THROUGH 130 ARE BASED ON THE FOLLOWING INFORMATION:


Consider a problem with the hypothesis test H 0 :   4 vs. H1 :   4 , and the following decision
x 4
rule: Reject H 0 if  1.645 .
0.08 / 25

125 Rewrite the decision rule explicitly in terms of x

ANSWER:
x 4
0.08 / 25
 
 1.645  x  4  1.645 0.08/ 25  4.026  Reject H 0 if x > 4.026

126. Compute the probability of Type II error and the power of the test for  = 4.06

ANSWER:
 4.026  4.06 
  P( X  4.026 |   4.06)  P  Z  
 0.08/ 25 
= P( Z  2.13)  0.50  0.4834  0.0166
Power = 1 -  = 1 – 0.0166 = 0.9834

127. Compute the probability of Type II error and the power of the test for  = 4.04

ANSWER:
 4.026  4.04 
  P( X  4.026 |   4.04)  P  Z  
 0.08/ 25 
= P(Z  0.88)  0.50  0.3106  0.1894
Power = 1 -  = 1 – 0.1894 = 0.8106

128. Compute the probability of Type II error and the power of the test for  = 4.01

ANSWER:
 4.026  4.01 
  P( X  4.026 |   4.01)  P  Z  
 0.08/ 25 
= P(Z  1.0)  0.50  0.3413  0.8413
Power = 1 -  = 1 – .8413 = 0.1587

273
Hypothesis Testing
129. Compute the probability of Type II error and the power of the test for  = 4.0

ANSWER:
 4.026  4.0 
  P( X  4.026 |   4.0)  P  Z  
 0.08/ 25 
= P( Z  1.63)  0.50  0.0.4484  0.9484
Power = 1 -  = 1 – .9484 = 0.0516

130. Use your answers to Questions 126, 127, 128, and 129 to graph the power function

ANSWER:

Power Function

1.2

0.8
Power

0.6

0.4

0.2

0
4 4.01 4.02 4.03 4.04 4.05 4.06 4.07
Mean

131. Explain carefully the distinction between null and alternative hypotheses.

ANSWER:
The null hypothesis is the statement that is assumed to be true unless there is sufficient
evidence to suggest that the null hypothesis can be rejected. The alternative hypothesis
is the statement that will be accepted if there is sufficient evidence to reject the null
hypothesis

QUESTIONS 132 THROUGH 136 ARE BASED ON THE FOLLOWING INFORMATION:


A wine producer claims that the proportion of its customers who cannot distinguish its product
from frozen grape juice is at most 0.12. The producer decides to test this null hypothesis against
the alternative that the true proportion is more than 0.12. The decision rule adopted is to reject
the null hypothesis if the sample proportion who cannot distinguish between these two flavors
exceeds 0.15.

132. If a random sample of 100 customers is chosen, what is the probability of a Type I error,
using this decision rule?

ANSWER:

274
Hypothesis Testing
H 0 : P  0.12 and H1 : P  0.12 . The decision rule is: Reject H 0 if p̂ > 0.15
 0.15  0.12 
  P( pˆ  0.15 | P  0.12)  P  Z  
 (0.12)(0.88) /100 

 P( Z  0.92)  0.50  0.0.3212  0.1788
P ( z > p – P0 / SQRT (p0(1-p0) / sqrt(n)

133. If a random sample of 400 customers is selected, what is the probability of a Type I error,
using this decision rule?

ANSWER:
 0.15  0.12 
  P( pˆ  0.15 | P  0.12)  P  Z  
 (0.12)(0.88) / 400 

 P( Z  1.85)  0.50  0.4678  0.0322
P ( z > p – P0 / SQRT (p0(1-p0) / sqrt(n)

134. Suppose that the true proportion of customers who cannot distinguish between these
flavors is 0.10. If a random sample of 100 customers is selected, what is the probability
of a Type II error?

ANSWER:
 0.15  0.20 
  P( pˆ  0.15 | P  0.20)  P  Z  
 (0.20)(0.80) /100 

= P( Z  1.25)  0.5  0.3944  0.1056
P ( z > p – P0 / SQRT (p0(1-p0) / sqrt(n)

135. Suppose that, instead of the decision rule, it is decided to reject the null hypothesis if the
sample proportion of customers who cannot distinguish between the two flavors exceeds
0.16. A random sample of 100 customers is selected. Without doing the calculations,
state whether the probability of a Type I error will be higher than, lower than, or the same
as in Question 132.

ANSWER:
Lower

136. Suppose that, instead of the decision rule, it is decided to reject the null hypothesis if the
sample proportion of customers who cannot distinguish between the two flavors exceeds
0.16. A random sample of 100 customers is selected. If the true proportion is 0.20, will
the probability of a Type II error be higher than, lower than, or the same as in Question
134?

ANSWER:
Higher

QUESTIONS 137 THROUGH 139 ARE BASED ON THE FOLLOWING INFORMATION:


An insurance company employs agents on a commission basis. It claims that, in their first year,
agents will earn a mean commission of at least $42,000 and that the population standard
deviation is no more than $6,800. A random sample of nine agents found, for commission in the

275
Hypothesis Testing
9 9

  x  x 
2
first year, xi  351,000 and i  338,000,000 . The population distribution can be
i 1 i 1
assumed to be normal.

137. What are the appropriate null and alternative hypotheses?

ANSWER:
H 0 :   42,000 and H1 :   42,000
H0 because at least 42,000

138. Calculate the sample mean and sample standard deviation

ANSWER:
9 9
x   xi / 9  351,000 / 9  39,000 and s   x  x  /8  338,000,000 /8  6,500
2
i
i 1 i 1

139. Test at the 5% level the null hypothesis in Question 137.

ANSWER:
Reject H 0 if t  t8,0.05  1.860
x  0 39,000  42,000
t   1.385 , therefore, fail to reject H 0 . We conclude that
s/ n 6,500 / 9
insurance company agents, in their first year, earn a mean commission of at least
$42,000.

QUESTIONS 140 THROUGH 143 ARE BASED ON THE FOLLOWING INFORMATION:


A certain pen has been designed so that true average writing lifetime under controlled conditions
(involving the use of a writing machine) is at least 12 hours. A random sample of 18 pens is
selected, the writing lifetime of each is determined, and a normal probability plot of the resulting
data supports the use of a one-sample t test.

140. What hypotheses should be tested if the investigator believes a priori that the design
specification has been satisfied?

ANSWER:
The appropriate hypotheses are H 0 :   12 vs. H1 :   12
H0:u = 12 because is says at least 12 hours and H1:u< 12 to contradict

141. What conclusion is appropriate if the hypotheses of Question 140 are tested, and the test
statistic t = -2.5, and   .05 ?

ANSWER:
-2.5 < - t17,.05  1.740, so we would reject H 0 . The data indicates that the pens do not
meet the design specifications.

142. What conclusion is appropriate if the hypotheses of Question 140 are tested, and the test
statistic t = -2, and   .01 ?

ANSWER:

276
Hypothesis Testing
-2.0 is not  t17,0.01  2.567 , so we would not reject H 0 . There is not enough evidence
to say that the pens don’t satisfy the design specifications.

143. What should be concluded if the hypotheses of Question 140 are tested and the test
statistic t = -3.25?

ANSWER:
Since t17,0.001  3.646  3.25  t17,.005  2.898 the p-value is between .001 and .005, which
gives strong evidence to support the alternative hypothesis.

277

Vous aimerez peut-être aussi