Question 1

Q1. Robin Inc. feared that the average company loss is running beyond £34,000. It initially
conducted a hypothesis test on a sample extracted from its database. The hypothesis
was formulated as H0 : average company loss ≤ £34,000; Ha : average company loss >
£34,000. The test resulted in favour of Robin Inc.’s loss not exceeding £34,000. Detailed
study of company accounts later revealed that the average company loss had run up to
£37,896. Which of the following errors were made during the hypothesis test?

A. Type I error
B. Type II error
C. Type III error
D. Type IV error

Question 2

Q2. An insurance agency wants to compare costs of automobile collision repair at a fran-
chise (F) repair shop with an independent (I) repair shop. The agency selects 30 cars at
random that are in need of various types of repair, and then takes each car to the franchise
repair shop and to the independent repair shop for a cost estimate. The difference in the
repair cost estimates will be determined for each car. A confidence interval will then be
used to determine if there is a statistically significant difference between the sample mean
repair costs at the 2 shops. The results from the samples that are provided to the agency
are as follows: x̄F − x̄I = £410; sd = £105
Use these results to determine a 99% confidence interval for µF − µI , and your conclusion.

A. (£360.64, £459.36) - Reject H0

B. (£360.64, £459.36) - Fail to reject H0
C. (£357.17, £462.83) - Reject H0
D. (£357.17, £462.83) - Fail to reject H0
E. (£357.28, £462.72) - Reject H0
F. (£357.28, £462.72) - Fail to reject H0
G. (£362.84, £457.16) - Reject H0
H. (£362.84, £457.16) - Fail to reject H0

Question 3

Q3. A manufacturer of light bulbs is considering a change in their supplier of tungsten for
use in the bulbs’ filaments. They believe this will result in bulbs that are “longer lasting”.
Currently, the lifetime of their bulbs, X , is a normally distributed random variable with popu-
lation mean µ = 750 hours. After using tungsten from the new supplier for a production run,
a random sample of 25 bulbs will be selected and the average lifetime of the bulbs will be

1
computed. This result will be used to determine if there is sufficient evidence to conclude
that the population mean lifetime of bulbs using this supplier is greater than 750 hours.
The results of the sample showed an average lifetime of 753.3 hours with a sample stan-
dard deviation of 7.5 hours. Which of the following statements is true? (Choose the most
appropriate response.)

A. The null hypothesis should be rejected for a level of significance of α = 0.01

B. The p-value of the test statistic is less than 0.025
C. The p-value of the test statistic is greater than 0.05
D. The null hypothesis should not be rejected (i.e. fail to reject) for a level of significance
of α = 0.05

Question 4

Q4. A soda bottling plant uses machines that initially filled bottles with an average of 12.26
fluid ounces. The machines are fairly old now and it is desired to know if this average has
changed. In a hypothesis test for this situation what would the alternative hypothesis be?

A. µ = 12.26
B. µ > 12.26
C. µ , 12.26
D. µ < 12.26
E. µ ≥ 12.26
F. µ ≤ 12.26

Question 5

Q5. The operations manager at a brewery that runs 24 hours per day has determined that
the amount of production time lost each day (24-hour period) due to equipment failure is a
normally distributed random variable, X . Historically, the population mean amount of lost
time has been assumed to be µ = 3.4 hours per day. The manager feels that due to aging
equipment, the population mean amount of lost time is now greater than 3.4 hours per day.
The manager decides to test the hypotheses H0 : µ ≤ 3.4 hours and Ha : µ > 3.4 hours
using the sample mean amount of time lost for a random sample of 36 days.
Over the next 2 months, the sample of 36 days results in an average amount of lost time
of x̄ = 3.92 hours per day with a sample standard deviation of s = 1.2 hours. Which of
the following conclusions can we draw for the test described above based on these sample
results? [note t0.05,35 = 1.690, t0.025,35 = 2.030, t0.01,35 = 2.438].

A. The null hypothesis should be rejected for any level of significance such that α > 0.01
B. The null hypothesis should not be rejected (i.e. fail to reject) for a level of significance
α = 0.01

2
 C. The null hypothesis is false and should be rejected
D. The p-value for the test statistics is greater than 0.025

Question 6

Q6. The operations manager at a brewery that runs 24 hours per day has determined that
the amount of production time lost each day (24-hour period) due to equipment failure is a
normally distributed random variable, X . Historically, the population mean amount of lost
time has been assumed to be µ = 3.4 hours per day.
The manager decides to upgrade the equipment and after 4 months, he will select another
random sample of 36 days and compute the average amount of lost time. In order to justify
the purchase of the new equipment, he is now looking for sufficient evidence to conclude
that the population mean amount of lost time is less than 3.4 hours per day. As such, the
hypotheses are H0 : µ ≥ 3.4 hours and Ha : µ < 3.4 hours, and he chooses to use a level
of significance of α = 0.025. If it is true that the population mean amount of lost time is
now µ = 2.21 hours per day and the standard deviation of the lost time is σ = 1.5 hours,
what is the probability that the manager will conclude that the population mean amount of
lost time is less than 3.4 hours? (Note: This is the correct conclusion to reach under these
conditions.)

A. 0.9750
B. 0.0026
C. 0.7704
D. 0.0250
E. 0.9974

Question 7

Q7. Which of the following would be an appropriate alternative hypothesis?

A. The mean of a population is equal to 55

B. The mean of a sample is equal to 55
C. The mean of a population is greater than 55
D. The mean of a population is greater than or equal 55

PLEASE TURN OVER

3
Questions 8 – 10

A sample of 300 peak electrical loads for consumers of Heysham Power are distributed
among three periods as follows:

Period Period definition Loads

1 2:00 AM - 10:00 AM 80
2 10:00 AM - 6:00 PM 90
3 6:00 PM - 2:00 AM 130

Table 1: Electrical loads at different times

To test the hypothesis that the loads are uniformly distributed among the periods, an analyst
decides to use a chi-squared hypothesis test.
Q8. Let p1 , p2 and p3 be the true proportions of loads in period 1, 2, and 3. Which
of the following should be his hypotheses statements?
A. H0 : p1 = p2 = p3 = 13 ; Ha : p1 , 13 or p2 , 13 or p3 , 13
B. H0 : p1 = p2 = p3 = 21 ; Ha : p1 , 12 or p2 , 12 or p3 , 12
C. H0 : p1 = 308 , p2 = 309 , p3 = 13
30
8
; Ha : p1 , 30 9
or p2 , 30 or p3 = 13
30
D. H0 : p1 = 0.267, p2 = 0.3, p3 = 0.433 ; Ha : p1 , 0.267 or p2 , 0.3 or p3 , 0.433
E. None of the above
Q9. Which of the following is the correct test statistic value?
A. 12
B. 13
C. 5.35
D. 6.63
E. None of the above
Q10. Which of the following is the correct conclusion at 5% significance level?
A. Test statistic is greater than χ25%,3 and hence the evidence from the sample
suggests that loads are not uniformly distributed
B. Test statistic is greater than χ25%,2 and hence the evidence from the sample
suggests that loads are not uniformly distributed
C. Test statistic is less than χ25%,3 and hence there is no evidence to say that loads
are not uniformly distributed
D. Test statistic is less than χ25%,2 and hence there is no evidence to say that loads
are not uniformly distributed
E. None of the above

END OF PAPER