BSc (Hons) Mathematics

Cohorts: BM/08/PT & BM/09/FT

Examinations for 2010– 2011 / Semester 1

MODULE: Linear Statistical Models

MODULE CODE: MATH 2132C

Duration: 2 Hours 30 Minutes

Instructions to Candidates:

1. Answer ALL the questions.

2. Questions may be answered in any order but your answers must

show the question number clearly.

3. Always start a new question on a fresh page.

4. All questions carry equal marks.

5. All workings should be clearly shown.

6. The Chi square Distribution, the Normal Distribution function and

F- Distribution tables are attached.

7. Total marks 100.

This question paper contains 4 questions and 7 pages.

Page 1 of 7
2010/S1
QUESTION 1: (25 MARKS)

1(a) A food products company is contemplating the introduction of a revolutionary

new product with new packaging to replace the existing product at much higher price
(A) or a moderate change in the composition of the existing product with a new
packaging at a small increase in price (B) or a small change in the composition of the
existing product except the word ‘New’ with a negligible increase in price (C).
The three possible states of nature or events are (i) high increase in sales (P), (ii) no
change in sales (Q), and (iii) decrease in sales (R). The marketing department of the
company worked out the payoffs in terms of yearly net profits for each of the
strategies of the events (expected sales). This is represented in the adjoining table.

States of Nature Payoffs (in ‘000 Rs)

Act A Act B Act C

P 70 50 30

Q 30 45 30

R 15 0 30

Which strategy should the executive concerned choose on the basis of

(i) Minimax Regret Criterion (2 marks)
(ii) Laplace Criterion (2 marks)
(iii) Hurwicz (assume that ) (2 marks)

Page 2 of 7
2010/S1
(b) Mr John buys a perishable commodity at Rs 5 each. The profit per unit is Rs 5.
This perishable commodity he can keep in his shop for a week and at the end of
each week the leftover are sold to a restaurant for Rs 3 each (at a loss of Rs 2
each). Mr John has given the record for the past 100 weeks for his weekly sales as
given below:

Weekly demand 1 2 3 4 5 6 7
Number of weeks 5 10 25 30 20 5 5

(i) Construct the conditional profit table. (5 marks)

(ii) Determine the optimum number of units of this commodity to order weekly, so
as to maximize his profit. (1 mark)
(iii) Compute the Expected Profit of Perfect Information (EPPI). (1 mark)
(iv) Compute the Expected Value of Perfect Information (EVPI). (1 mark)
(v) Interpret the value of EVPI. (1 mark)

(c) A businessman has two independent investments A and B available to him but he
lacks the capital to undertake both of them simultaneously. He can choose to take A
first and then stop, or if A is successful then take B, or vice versa. The probability of
success for A is 0.65 while for B it is 0.45. Both investments require an initial capital
outlay of Rs 2,500.; and both return, nothing if the venture is unsuccessful.
Successful completion of A will return Rs 3,500 (over cost) and successful
completion of B will return Rs 5500 (over cost). Draw and evaluate the decision tree
by the roll back technique and determine the best strategy. (10 marks)

Page 3 of 7
2010/S1
QUESTION 2: (25 MARKS)

2(a) An old film is treated with a chemical in order to improve the contrast.
Preliminary tests on nine samples drawn from a segment of the film produced the
following results.

Sample A B C D E F G H I
x 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
y 49 60 66 62 72 64 89 90 96

The quantity is a measure of the amount of chemical applied, and is the contrast
index, which takes values between 0 (no contrast) and 100 (maximum contrast).
(i) Plot a scatter diagram to illustrate the data. (3 marks)
(ii) It is subsequently discovered that one of the samples of film was damaged and
produced an incorrect result. State which sample you think this was. (1 mark)

In all subsequent calculations this incorrect sample is ignored. The remaining data
can be summarized as follows:
, , ∑ , , ,

(iii) Calculate the product moment correlation coefficient. (5 marks)

(iv) State, with a reason, whether it is sensible to conclude from your answer to
part (iii) that and are linearly related. (2 marks)

(v) The line of regression of on has equation . Calculate the values

of and , each correct to 3 significant figures. (5 marks)

(vi) Use your regression equation to estimate what the contrast index corresponding
to the damaged piece of film would have been if the piece had been undamaged.
(2 marks)
(vii) State, with a reason, whether it would be sensible to use your regression
equation to estimate the contrast index when the quantity of chemical applied to the
film is zero. (2 marks)

Page 4 of 7
2010/S1
2(b) Seven army recruits (A, B, …,G) were given two separate aptitude tests. Their
orders of merit in each test were as follows:

Order of merit 1st 2nd 3rd 4th 5th 6th 7th

1st test G F A D B C E
2nd test D F E B G C A

Find the Spearman’s coefficient of rank correlation between the two orders and
comment briefly on the correlation obtained. (5 marks)

Page 5 of 7
2010/S1
QUESTION 3: (25 MARKS)

3(a) A farmer applies three types of fertilizers on 4 separate plots. The figures on
yield per acre are tabulated below.

Yield
Fertilisers Plot A Plot B Plot C Plot D Total
Nitrogen 6 4 8 6 24
Potash 7 6 6 9 28
Phosphates 8 5 10 9 32
Total 21 15 24 24 84

Perform an analysis of variance to test that

(i) The three fertilizers make any material difference in yields. (1 mark)
(ii) The plots are materially different in fertility. (1 mark)

You should construct an appropriate ANOVA table and state your conclusion clearly.
(13 marks)

3(b) Professor Newton is interested in determining whether his brightest students

(those making the best grades) tend to turn in their tests earlier (because they can
recall the material faster) or later (because they take longer to write down all they
know) than the others in the class. For a particular physics test, he observes that the
students make the following grades in order of turning their tests in:

Order Grades
1-10 94 70 85 89 92 98 63 88 74 85
11-20 69 90 57 86 79 72 80 93 66 74
21-30 50 55 47 59 68 63 89 51 90 88

If Professor Newton counts those making a grade of 90 and above as his brightest
students, then at a 5 percent level of significance, can he conclude that the brightest
students turned their tests in randomly? (10 marks)

Page 6 of 7
2010/S1
QUESTION 4: (25 MARKS)

4. (a) Melisa’s Boutique has three small locations. Melisa keeps a daily record for
each location of the number of customers who actually make a purchase. A sample
of those data follows. Using the Kruskal-Wallis, test, at the 0.05 level of significance
that her stores have the same number of customers who buy. (11 marks)

Trianon 99 64 101 85 79 88 97 95 90 100

Caudan 83 102 125 61 91 96 94 89 93 75
Grand Baie 89 98 56 105 87 90 87 101 76 89

(b) A large hospital hires most of its nurses from the universities in the area. Over
the last year, they have been giving a test to the newly graduated nurses entering
the hospital to determine which school, if either, seems to educate nurses better.
Based on the following scores (out of 100 possible points), help the personnel office
of the hospital determine whether the schools differ in quality. Use the Mann-
Whitney U test with a 10 percent level of significance. (14 marks)

Test Score
School A 97 69 73 84 76 92 90 88 84 87 93
School B 88 99 65 69 97 84 85 89 91 90 87 91 72

***END OF QUESTION PAPER***

Page 7 of 7
2010/S1