HARARE INSTITUTE OF TECHNOLOGY

SCHOOL OF INFORMATION SCIENCE AND TECHNOLOGY

BTech degree in Computer Science

BTech degree in Information Technology
BTech degree in Information Security and Assurance.

Applied Statistics: ICS2205

Date: June 2022 Time: 3 hours

INTRUCTIONS TO CANDIDATES

1 This question paper consists of four (5) printed pages including the cover page.

2 A formula sheet should accompany this question paper, if it is not provided ask for it.

3 A statistical table should be provided.

4 Answer ALL questions in section A and any THREE questions in section B.

ADDITIONAL MATERIALS

1. Formula Sheet.

2. Statistical Table.

page 1 of 5
 ICS2205

SECTION A (40 marks)

Candidates must attempt ALL questions being careful to number them A1 to A4.

A1. Two flow meters, A and B, are used to measure the flow rate of brine in a potash pro-
cessing plant. The two meters are identical in design and calibration and are mounted
on two adjacent pipes, A on pipe 1 and B on pipe 2. On a certain day, the following
flow rates (in m3 /sec) were observed at 10-minute intervals from 1:00 p.m. to 2:00
p.m.
Meter A 1.7 1.6 1.5 1.4 1.5 1.6 1.7
Meter B 2 1.8 1.6 1.3 1.6 1.7 1.9

(a) Using R, write down commands necessary to test if the flow in pipe 2 is signifi-
cantly different from the flow in pipe 1 at the 5% level of significance? [4]
(b) Based on the data above, interpret the following R output: focusing on hypothesis,
significance of the test and the confidence interval. [4]
Paired t-test

data: meterB and meterA

t = 2.7136, df = 6, p-value = 0.03494
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
0.01263588 0.24450698
sample estimates:
mean of the differences
0.1285714

A2. (a) The probability of failure of a mechanical device as a function of time is given by
the following pdf:
−3t

 3e , t ≥ 0

f (t) =


0, t<0
(i) Find the mean lifetime of the device. [2]
(ii) Find the variance of the lifetime of the device. [3]
(iii) What is the probability that the device will fail within one standard deviation
of its mean lifetime? [4]
(b) It is known that of the articles produced by a factory, 10% come from Machine A,
40% from Machine B, and 50% from Machine C. The percentages of satisfactory
articles among those produced are 97% for A, 90% for B and 95% for C. An
article is chosen at random. Assuming that the article is satisfactory, what is the
probability that it was produced by Machine A? [5]

page 2 of 5
 ICS2205

A3. (a) Consider the following data on the crack length (µm) as a result of constant load
stress corrosion tests on smooth bar tensile specimens for a fixed length of time.

16.1 9.6 24.9 20.4 12.7 21.2 30.2 25.8

18.5 10.3 25.3 14.0 27.1 45.0 23.3 24.2

Construct a box plot of the data and identify any outlying observations. [5]
(b) Consider the following measurements (in the same units) from three machines:

Machine x̄ s2
A 5.92 0.0002
B 10.035 2.56
C 8.22 0.0153

Advise management on which machine is more precise. [4]

A4. A company has set up a production line for cans of carrots. The numbers of breakdowns
on the production line over 49 shifts are summarized as follows:

Number of breakdowns in one shift Number of shifts

0 18
1 12
2 11
≥3 8

Is this distribution significantly different from a Poisson distribution? Use the 5% level
of significance. [9]

SECTION B (60 marks)

Candidates must attempt THREE questions being careful to number them B5 to B8.

B5. (a) A time study engineer is assigned the problem of two different work sequences
in a garment factory for measuring the shear strength of polyester fibres. To
collect some data, he randomly divides 12 workers into two groups. The first
group measures the shear strength using work sequence A, the second group,
work sequence B. The data recorded were the 12 completion times.

Work Sequence A Work Sequence B

220 247
235 223
214 215
197 219
206 207
214 236

page 3 of 5
 ICS2205

(i) Test whether the difference in completion times is significantly different. Use
α = 0.05. [7]
(ii) Construct a 95% confidence interval in mean difference of completion times. [3]
(b) A recent investigation looked at nightmare frequencies for a sample of 160 men and
192 women. Each subject was asked whether he (or she) experienced nightmares
often (at least once a month) or seldom (less than once a month). The findings
are summarised below:
Men Women Total
Nightmares often 55 60 115
Nightmares seldom 105 132 237
Total 160 192 352

Let Pm and Pw denote true proportions of men having nightmares often and
women having nightmares often respectively.
(i) Test the hypothesis H0 : Pm = Pw versus H1 : Pm ̸= Pw . Use α = 0.05. [7]
(ii) Construct a 95% confidence interval for the difference in proportions. [3]

B6. A paper in the Journal of the Association of Asphalt Paving Technologists (Vol. 59,
1990) describes an experiment to determine the effect of air voids on percentage re-
tained strength of asphalt. For purposes of the experiment, air voids are controlled at
three levels: low (2-4%), medium (4-6%) and high (6-8%). The data are shown below:

Air voids Retained Strength(%)

Low 106 90 103 90 79 88 92 95
Medium 80 69 94 91 70 83 87 83
High 78 80 62 69 76 85 69 85

(a) Explaining all parameters used, propose an appropriate statistical model for this
data. [5]
(b) Do the different levels of air voids significantly affect mean retained strength? Use
α = 0.05. [10]
(c) Using Fisher’s LSD, compare the mean retained strength between the low and
high air voids. Use α = 0.05. [5]

B7. (a) Consider a simple linear regression model of the form yi = β0 + β1 xi + ϵi . Let ŷi
be the fitted model and ri be the ith residual. Prove the following:
n
X
(i) ri = 0. [3]
i=1
Xn
(ii) xi ri = 0. [3]
i=1

page 4 of 5
 ICS2205

(b) An article in Wear (Vol. 152, 1992, pp. 171-181) presents data on fretting wear
of mild steel and oil viscosity. x= oil viscosity, and y= wear volume (10−4 cubic
millimeters).

x 1.6 9.4 15.5 20 22 35.5 43 40.5 33

y 240 181 193 155 172 110 113 75 94

(i) Use the correlation coefficient to establish the nature of the relationship be-
tween oil viscosity and wear. [3]
(ii) Fit a simple linear regression model of the form yi = β0 + β1 xi + ϵi . [3]
(iii) Predict fretting wear when viscosity x=30. [1]
(iv) Test the hypothesis H0 : β1 = 0 versus H1 : β1 ̸= 0. Use α = 0.05. [6]
(v) What percentage of fretting wear is being explained by the model? [1]

B8. (a) The operations manager of a company that manufactures tyres wants to determine
whether there are any differences in the quality of work among the three daily
shifts. She randomly selects 496 tyres and carefully inspects them. Each tyre is
either classified as perfect, satisfactory, or defective, and the shift that produced
it is also recorded. The two categorical variables of interest are the shift and
condition of the tyre produced.

Condition of the tyre

Shift Perfect Satisfactory Defective
Shift 1 106 124 1
Shift 2 67 85 1
Shift 3 37 72 3

Does the data provide sufficient evidence at the 5% significance level to infer that
there are differences in quality among the three shifts? [10]
(b) Three chemical sprays for killing flies are tested, and the percentage of flies killed
was recorded as follows:
Brand A 72 65 67 75 62 73
Brand B 55 59 68 70 53 50
Brand C 64 74 61 58 51 69

Using the Kruskal Wallis procedure, test whether these data provide sufficient
evidence to indicate a difference in dead flies. [10]

END OF QUESTION PAPER

page 5 of 5