MATH 3

EXAM
September 2022

Full Name

Student ID

Project Group Number

Signature

Please answer all questions and provide detailed working. The total mark
is 100. The test is worth 100% of the final grade.

Time: 120 minutes

Question Marks achieved Out of

1 20
2 20
3 10
4 20
5 10
6 10
Total 100
Question 1 (20 marks) An aircraft can seat 220 passengers and each of the passengers
booked on the flight has a probability of 0.9 of actually arriving at the gate to board the
plane, independent of the other passengers.

(a) Suppose the airline books 235 passengers on the flight. What is the probability that
there will be insufficient seats to accommodate all of the passengers who wish to
board the plane?

(b) If the airline wants to be 75% confident that there will be no more than 220 passengers
who wish to board the plane, how many passengers can be booked on the flight?

1
Question 2 (total 20 marks) A company has installed 3288 flow meters throughout
an extensive sewer system. Unknown to the company, 592 of these meters are operating
outside acceptable tolerance limits, whereas the other 2696 meters are operating satisfac-
torily. The company decides to estimate the unknown proportion p of the meters that are
operating outside acceptable tolerance limits based on the inspection of a random sample
of 20 meters.

(a) What is the probability that the company’s estimate of p will be within 0.1 of the
correct value?

(b) Suppose that 2012 of the meters are easily accessible, whereas the other 1276 meters
are not easily accessible. In addition, suppose that only 184 of the easily accessible
meters are operating outside acceptable tolerance limits. If the company’s sample
of 20 meters is biased due to the fact that the meters were randomly chosen from
the subset of easily accessible meters, what is the probability that the company’s
estimate of p will be within 0.1 of the correct value?

2
Question 3 (10 marks) A boot manufacturer is testing the quality of leather provided
by a potential supplier. The manufacturer wants to construct a two-sided confidence
interval with a confidence level of 95% that has a length no larger than 0.1, and from
previous experience it is believed that the variability in the leather is such that the standard
deviation is no larger than 0.2031. What sample size would you recommend?

3
Question 4 (20 marks) Twelve samples of a metal alloy are tested. The flexibility
measurements had a sample average of 732.9 and a sample standard deviation of 12.5.

(a) Is there sufficient evidence to conclude that the flexibility of this kind of metal alloy
is smaller than 750? Use an appropriate hypothesis test to investigate this question.

(b) Construct a 99% confidence interval that provides an upper bound on the flexibility
of this kind of metal alloy.

4
Question 5 (15 marks) In an unpaired two-sample problem an experimenter observes
n = 10, x = 7.76, sx = 1.07 from population A and m = 9, y = 6.88, sy = 0.62 from
population B.

(a) Construct a 99% one-sided confidence interval µA − µB ∈ (c, ∞) without assuming

equal population variances.

(b) Does the value of c increase or decrease if a confidence level 95% is used?

5
(c) Consider a one-sided hypothesis test of

H0 : µA ≤ µB versus HA : µA > µB

without assuming equal population variances. Does a size α = 0.01 test accept or
reject the null hypothesis? Calculate the exact p-value.

6
Question 6 (15 marks) The dimensions of 3877 manufactured parts were examined
and 445 were found to have a length outside a specified tolerance range.

(a) Conduct a hypothesis test to investigate whether there is sufficient evidence to con-
clude that the probability of a part having a length outside the tolerance range is
larger than 10%.

(b) Construct a one-sided 99% confidence interval that provides a lower bound on the
probability of a part having a length outside the tolerance range.

7
(c) It was also found that out of the 445 parts that had a length outside the specified
tolerance range, 25 also had a width outside a specified tolerance range. Furthermore,
out of the 3432 parts that had a length inside the specified tolerance range, it was
found that 161 had a width outside the specified tolerance range.
Use the Pearson chi-square statistic to test whether the acceptability of the length
and the acceptability of the width of the parts are related to each other, or whether
these two factors can be considered to be independent of each other.

8
Full name .................................................................... Signature ........................................................

BLANK PAGE FOR EXTRA WORKING (Any work here will NOT be marked.)