LUBS267001

©UNIVERSITY OF LEEDS
January 2013

Examination for the degree of

BA and BSc

STATISTICS FOR BUSINESS AND ECONOMICS 2

ANSWERS
QUESTION ANSWER QUESTION ANSWER
1 D 31 C
2 C 32 D
3 A 33 A
4 A 34 B
5 D 35 A
6 E 36 A
7 E 37 D
8 B 38 B
9 A 39 C
10 D 40 B
11 D 41 D
12 B 42 C
13 E 43 E
14 E 44 D
15 C 45 A
16 D 46 E
17 D 47 C
18 E 48 B
19 C 49 E
20 D 50 B
21 B 51 A
22 C 52 B
23 E 53 C
24 A 54 C
25 B 55 C
26 D 56 B
27 B 57 A
28 D 58 D
29 E 59 C
30 C 60 A

2
1. If X and Y are discrete random variables and f ( x, y ) is their joint probability
distribution, the general definition of marginal distribution of X is

f ( x ) = ∑ f ( x, y ) for x = x1 , x 2 , ..., x n
y

In our case, f ( x, y ) =
1
128
(x 2 + y 2 ), x = −1, 2, 3, y = −2, 0, 2, 4 . So we have

(x + y2 )
4
f (x ) =
1 2
∑
y = −2 128

ANSWER: D

(x + y2 ) =
1 ⎛ 4 2 ⎞
[ ]
4 4
f (x ) = 4 x 2 + (− 2 ) + 02 + 22 + 42 =
1 2 1
2. ∑ 128
⎜ ∑ x + ∑ y2 ⎟ =
⎜
128 ⎝ y = −2 ⎟
2

y = −2 y = −2 ⎠ 128

=
1
128
(4 x 2 + 4 + 0 + 4 + 16) =
1
128
(4 x 2 + 24) =
4 2
128
(x + 6) = (x + 6)
1 2
32

Therefore, the marginal distribution of X is

f (x ) = (x + 6),
1 2
32
x = −1, 2, 3

ANSWER: C

3. If X and Y are discrete random variables and f ( x, y ) is their joint probability

distribution, the general definition of marginal distribution of Y is

f ( y ) = ∑ f ( x, y ) for y = y1 , y2 , ..., ym
x

In our case, f ( x, y ) =
1
128
(x 2 + y 2 ), x = −1, 2, 3, y = −2, 0, 2, 4 . So we have

∑ 128 (x + y2 )
3
f (y) =
1 2

x = −1

ANSWER: A

3
 ∑ 128 (x + y2 ) =
1 ⎛ 3 2 ⎞
[ ]
3 3
f (y) = (− 1)2 + 22 + 32 + 3 y 2 =
1 1
4. 2
⎜ ∑ x + ∑ y 2 ⎟⎟ =
x = −1 128 ⎝ x = −1 x = −1 ⎠ 128

=
1
128
(1 + 4 + 9 + 3y2 ) =
1
128
(14 + 3 y 2 )

Therefore, the marginal distribution of Y is

f (y) =
1
128
(3 y 2 + 14), y = −2, 0, 2, 4

ANSWER: A

5. The formula for the required conditional distribution is

1
f ( X = −1, y ) 128
(− 1) + y 2
2 1
[ (1 + y 2 ) ]
f ( y X = −1) = = = 128 =
f ( X = −1) 1
32
(− 1)2 + 6 [ 1
32
(7) ]

=
32 1
⋅
7 128
(
1 + y 2 ) = ⋅ (1 + y 2 ) =
1 1
7 4
1
28
(
1 + y2 )

Therefore, the conditional probability distribution of Y , the monthly profit of the

shopping centre store, when X = −1 – i.e., when the inner city store makes a loss of
£ 1,000 – is

f ( y X = −1) =
1 2
28
(y + 1), y = −2, 0, 2, 4

ANSWER: D

E (Y X = −1) = y ⋅ f ( y X = −1) = (y + 1) =
4 4
1 2
6. ∑
y = −2
∑
y = −2
y⋅
28

1 ⎛ 4 ⎞
1 4
( y 3 + y ) = ⎜⎜ ∑ y 3 + ∑ y ⎟⎟ =
4
= ∑
28 y = −2 28 ⎝ y = −2 y = −2 ⎠

=
1
28
[
(− 2 )3 + 03 + 23 + 43 + (− 2) + 0 + 2 + 4 = ]

=
1
( − 8 + 0 + 8 + 64 − 2 + 0 + 2 + 4 ) = 1 (64 + 4) =
28 28

4
 1
(68) = 17 = 2.42857
28 7

Since X and Y are measured in thousand pounds, the conditional expected monthly
profit of the shopping centre store when the inner city store makes a loss of £ 1,000 is
therefore £ 2,428.57 .

ANSWER: E

7. The random variables X and Y are independent if f ( x ) f ( y ) = f (x, y ) . In terms of

conditional probability, this implies:

f (x, y ) f ( x ) f ( y )
f (y x ) = = = f (y) ∀ x, y over the whole sets of their
f (x ) f (x )
possible values

ANSWER: E

8. We can use the following property of a probability density function ( pdf ):

∫ f (x )dx = 1,
x=a
a<x<b

i.e., the sum of all probabilities is equal to 1.

The flight-time random variable X is uniformly distributed between 1 hour, 50

minutes and 2 hours, 30 minutes. The uniform pdf of X expressed in minutes is

f (x ) =
1
c
135

Note that: 1 hour, 50 minutes = ( 60 + 50 ) minutes = 110 minutes

2 hours, 30 minutes = ( 60 ⋅ 2 + 30 ) minutes = 150 minutes

Therefore, the range of definition a < x < b for X expressed in minutes is

110 < x < 150

So in our case we have

150
⎛ 1 ⎞ ⎡ ⎤
150
c (150 − 110 ) = c (40 ) =
1 1 1 8
∫
x =110
⎜ c ⎟ dx =
⎝ 135 ⎠ 135
c⎢ x ⎥ =
⎣ ⎦110 135 135 27
c =1

5
 8 27
∴ c =1 ⇒ c =
27 8

Alternatively, for a uniform pdf we have

f (x ) =
1
, a<x<b
b−a

The range of definition a < x < b for X expressed in minutes is

110 < x < 150

So for the uniform pdf of X expressed in minutes we have

f (x ) =
1 1 1 1
= = = c
b − a 150 − 110 40 135

1 1 1 135 27
∴ c= ⇒ c= ⋅ =
135 40 40 1 8

f (x ) =
1
The uniform pdf is therefore: , 110 < x < 150
40

ANSWER: B

9. The general definition of cumulative probability distribution for a continuous random

variable is

x
F (x ) = ∫ f (t )dt , a< x<b
t =a

In our case

x
⎛ 1 ⎞
F (x ) = ∫ ⎜ c ⎟ dt
t =110 ⎝
135 ⎠

ANSWER: A

x
⎞ 1 ⎛ 27 ⎞ ⎛ 1 ⎞
( 1 )dt = 1 ⎡⎢ t ⎤⎥ =
x x x x
⎛ 1
F (x ) =
1
10. ∫ ⎜
t =110 ⎝
135
c ⎟ dt = ∫
⎠ t =110
⎜ ⎟ dt = ∫ ⎜ ⎟ dt =
135 ⎝ 8 ⎠ t =110 ⎝
40 ⎠ ∫
40 t =110 40 ⎣ ⎦110

=
1
(x − 110) = 1 x − 110 = 1 x − 11
40 40 40 40 4

6
 The cumulative probability distribution is therefore:

F (x ) =
1 11
x− , 110 < x < 150
40 4

ANSWER: D

F (x ) =
1 11
11. x− , 110 < x < 150
40 4

equation of a linear function.

x = 110 → F ( x ) = (110) − 11 = 11 − 11 = 0
1
40 4 4 4

x = 150 → F ( x ) = (150) − 11 = 15 − 11 = 4 = 1
1
40 4 4 4 4

F (x )

x
0 110 150

ANSWER: D

12. Ryanair quotes a flight time of 2 hours, 15 minutes, which expressed in minutes is

2 hour, 15 minutes = ( 60 ⋅ 2 + 15 ) minutes = 135 minutes

Therefore a flight will be more than 10 minutes late if it lasts more than ( 135 + 10 )
minutes = 145 minutes.

The required probability is thus

P ( X > 145), 110 < x < 150

7
 We can use the property of the cpd F ( x ) :

P (a < X < b ) = F (b ) − F (a )

In our case we have

P ( X > 145) = P (145 < X < 150 ) = F (150 ) − F (145) =

⎡1 11⎤ ⎡ 1 11⎤ 150 11 145 11 5 1

= ⎢ (150 ) − ⎥ − ⎢ (145) − ⎥ = − − + = =
⎣ 40 4 ⎦ ⎣ 40 4 ⎦ 40 4 40 4 40 8

Alternatively, we can use the property of the pdf f ( x ) :

b
P (a ≤ X ≤ b ) = ∫ f ( x ) dx
a

In our case we have

⎛ 1 ⎞
150 150 150
P ( X > 145) = P (145 < X < 150 ) = ∫ f (x )dx = ∫ ( 1 )dx =
1
40 x =∫145
⎜ ⎟ dx =
x =145 x =145 ⎝ 40 ⎠

150
1 ⎡ ⎤
= ⎢ x⎥ =
1
(150 − 145) = 1 (5) = 1
40 ⎣ ⎦145 40 40 8

ANSWER: B

13. If Y = d + e X is a linear combination of X – where d and e are constants – the

expected value of Y , E (Y ) , is given by the formula for the expectation of a linear
combination of a random variable:

E (Y ) = E (d + e X ) = d + e E ( X )

Since Y = 75 + 2 X , in our case d = 75 and e = 2 .

From the properties of the uniform distribution, we know that

a+b
f (x ) = E(X ) =
1
for , a < x < b:
b−a 2

In our case

110 + 150 260

f (x ) = E(X ) =
1
for , 110 < x < 150 : = = 130
40 2 2

8
 So we have:

E (Y ) = 75 + 2 (130 ) = 75 + 260 = 335

i.e., expected fuel costs are £335.00.

ANSWER: E

14. If Y = d + e X is a linear combination of X – where d and e are constants – the

variance of Y , var (Y ) , is given by the formula for the variance of a linear
combination of a random variable:

var (Y ) = var (d + e X ) = e 2 var ( X )

Since Y = 75 + 2 X , in our case d = 75 and e = 2 .

From the properties of the uniform distribution, we know that

for f (x ) =
1
, a < x < b: var ( X ) =
(b − a ) 2
b−a 12

In our case

for f ( x ) = , 110 < x < 150 :

1
var ( X ) =
(150 − 110 )
2
=
402
40 12 12

So for the variance of fuel costs we have:

⎛ 40 2 ⎞
var (Y ) = 2 ⎜⎜
2
⎟⎟
⎝ 12 ⎠

Taking the square root of both sides, we have the standard deviation of fuel costs:

⎛ 40 ⎞ 80
SD (Y ) = 2 ⎜ ⎟= = 23.09
⎝ 12 ⎠ 12

i.e., the standard deviation of fuel costs is £23.09.

ANSWER: E

15. If X and Y are continuous random variables and f ( x, y ) is their joint probability
distribution, the general definition of marginal distribution of X is

9
 d
f (x ) = ∫ f (x, y ) dy for a< x<b
y =c

f ( x, y ) = (x y + x + y + 1), − 0.05 < x < 0.15 , − 0.10 < y < 0.10 .

500
In our case,
21
So we have

0.10
f (x ) = (x y + x + y + 1) dy
500
∫
y = −0.10
21

ANSWER: C

0.10
f (x ) = (x y + x + y + 1) dy
500
16. ∫
y = −0.10
21

0.10 0.10
= ∫
500
(x + 1) ( y + 1) dy = 500 (x + 1) ∫ ( y + 1) dy =
y = −0.10
21 21 y = −0.10

⎡⎛ ⎞⎤
0.10
⎡ 2 ⎤ ⎞ ⎛
(x + 1) ⎢ y + y ⎥ = 500 (x + 1) ⎢⎜⎜ 0.10 + 0.10 ⎟⎟ − ⎜⎜ (−0.10) + (−0.10) ⎟⎟⎥ =
2 2
500
=
21 ⎣2 ⎦ − 0.10 21 ⎣⎝ 2 ⎠ ⎝ 2 ⎠⎦

⎛ 2 2
⎞
=
500
(x + 1) ⎜⎜ 0.10 + 0.10 − 0.10 + 0.10 ⎟⎟ = 500 (x + 1) (0.20) =
21 ⎝ 2 2 ⎠ 21

=
500
(x + 1) ⎛⎜ 1 ⎞⎟ = 100 (x + 1)
21 ⎝ 5 ⎠ 21

Therefore, the marginal distribution of X is

f (x ) = (x + 1),
100
− 0.05 < x < 0.15
21

ANSWER: D

0.15
17. E(X ) = ∫ x f ( x ) dx =
x = −0.05

0.15
100 ⎡ x 3 x 2 ⎤
∫ (x + x )dx = 21 ⎢⎣ 3 + 2 ⎥⎦ =
0.15 0.15
= ∫ x⋅
100
(x + 1)dx = 100 2

x = −0.05
21 21 x = −0.05 − 0.05

10
 100 ⎡⎛ 0.153 0.152 ⎞ ⎛ ( −0.05)3 ( −0.05)2 ⎞⎤
= ⎢⎜ + ⎟−⎜ + ⎟⎟⎥ =
21 ⎣⎜⎝ 3 2 ⎟⎠ ⎜⎝ 3 2 ⎠⎦

100 ⎡⎛ 0.003375 0.0225 ⎞ ⎛ 0.000125 0.0025 ⎞⎤

= ⎜ + ⎟ − ⎜− + ⎟ =
21 ⎢⎣⎝ 3 2 ⎠ ⎝ 3 2 ⎠⎥⎦

100 ⎛ 0.003375 0.0225 0.000125 0.0025 ⎞

= ⎜ + + − ⎟=
21 ⎝ 3 2 3 2 ⎠

100 ⎛ 0.0035 0.02 ⎞ 100 ⎛ 0.0035 ⎞ 100 ⎛ 0.0035 + 0.03 ⎞

= ⎜ + ⎟= ⎜ + 0.01⎟ = ⎜ ⎟=
21 ⎝ 3 2 ⎠ 21 ⎝ 3 ⎠ 21 ⎝ 3 ⎠

100 ⎛ 0.0335 ⎞ 100 ⎛ 1 ⎞ ⎛ 335 ⎞ 1 ⎛ 1 ⎞ ⎛ 67 ⎞ 67

= ⎜ ⎟= ⎜ ⎟⎜ ⎟ = ⎜ ⎟⎜ ⎟ = = 0.05317
21 ⎝ 3 ⎠ 21 ⎝ 3 ⎠ ⎝ 10000 ⎠ 21 ⎝ 3 ⎠ ⎝ 20 ⎠ 1260

ANSWER: D

E (X 2 ) =
0.15
18. ∫ x 2 f ( x ) dx =
x = −0.05

0.15
100 ⎡ x 4 x 3 ⎤
( )
0.15 0.15
= ∫ x ⋅
2 100
(x + 1)dx = 100 ∫
3
+2
= + ⎥ =
21 ⎢⎣ 4
x x dx
x = −0.05
21 21 x = −0.05
3 ⎦ − 0.05

100 ⎡⎛ 0.154 0.153 ⎞ ⎛ ( −0.05) 4 ( −0.05)3 ⎞⎤

= ⎢⎜ + ⎟−⎜ + ⎟⎟⎥ =
21 ⎣⎜⎝ 4 3 ⎟⎠ ⎜⎝ 4 3 ⎠⎦

100 ⎡⎛ 0.00050625 0.003375 ⎞ ⎛ 0.00000625 0.000125 ⎞⎤

= ⎜ + ⎟−⎜ − ⎟⎥ =
21 ⎢⎣⎝ 4 3 ⎠ ⎝ 4 3 ⎠⎦

100 ⎛ 0.00050625 0.003375 0.00000625 0.000125 ⎞

= ⎜ + − + ⎟=
21 ⎝ 4 3 4 3 ⎠

100 ⎛ 0.0005 0.0035 ⎞ 100 ⎛ 0.0015 + 0.014 ⎞

= ⎜ + ⎟= ⎜ ⎟=
21 ⎝ 4 3 ⎠ 21 ⎝ 12 ⎠

100 ⎛ 0.0155 ⎞ 100 ⎛ 1 ⎞ ⎛ 155 ⎞ 1 ⎛ 1 ⎞ ⎛ 31 ⎞ 31

= ⎜ ⎟= ⎜ ⎟⎜ ⎟ = ⎜ ⎟⎜ ⎟ = = 0.00615
21 ⎝ 12 ⎠ 21 ⎝ 12 ⎠ ⎝ 10000 ⎠ 21 ⎝ 12 ⎠ ⎝ 20 ⎠ 5040

ANSWER: E

11
19. ( )
var( X ) = E X 2 − [E ( X )] =
2

= 0.00615 − 0.053172 = 0.00615 − 0.00283 = 0.00332

ANSWER: C

20. If X and Y are continuous random variables and f ( x, y ) is their joint probability
distribution, the general definition of marginal distribution of Y is

b
f (y) = ∫ f (x, y )dx for c< y<d
x =a

f ( x, y ) = (x y + x + y + 1), − 0.05 < x < 0.15 , − 0.10 < y < 0.10 .

500
In our case,
21
So we have

0.15
f (y) = (x y + x + y + 1)dx
500
∫
x = −0.05
21

ANSWER: D

0.15
f (y) = (x y + x + y + 1)dx
500
21. ∫
x = −0.05
21

0.15 0.15
= ∫
500
(x + 1) ( y + 1)dx = 500 ( y + 1) ∫ (x + 1)dx =
x = −0.05
21 21 x = −0.05

⎡⎛ ⎞⎤
0.15
⎡ 2 ⎤ ⎞ ⎛
( y + 1) ⎢ x + x ⎥ = 500 ( y + 1) ⎢⎜⎜ 0.15 + 0.15⎟⎟ − ⎜⎜ (−0.05) + (−0.05) ⎟⎟⎥ =
2 2
500
=
21 ⎣2 ⎦ − 0.05 21 ⎣⎝ 2 ⎠ ⎝ 2 ⎠⎦

=
500
( y + 1) ⎡⎢⎛⎜ 0.0225 + 0.15⎞⎟ − ⎛⎜ 0.0025 − 0.05⎞⎟⎤⎥ =
21 ⎣⎝ 2 ⎠ ⎝ 2 ⎠⎦

=
500
( y + 1) ⎛⎜ 0.0225 + 0.15 − 0.0025 + 0.05⎞⎟ = 500 ( y + 1) ⎛⎜ 0.02 + 0.20 ⎞⎟ =
21 ⎝ 2 2 ⎠ 21 ⎝ 2 ⎠

=
500
( y + 1) (0.01 + 0.20) = 500 ( y + 1) (0.21) = 500 ( y + 1) ⎛⎜ 21 ⎞⎟ = 5 ( y + 1)
21 21 21 ⎝ 100 ⎠

Therefore, the marginal distribution of Y is

12
 f ( y ) = 5 ( y + 1), − 0.10 < y < 0.10

ANSWER: B

0.10
22. E (Y ) = ∫ y f ( y ) dy =
y = −0.10

0.10
⎡ y3 y 2 ⎤
= ∫ y ⋅ 5 ( y + 1) dy = 5 ∫ (y + y )dy = 5 ⎢ + ⎥
0.10 0.10
2
=
y = −0.10 y = −0.10 ⎣3 2 ⎦ − 0.10

⎡⎛ 0.103 0.10 2 ⎞ ⎛ ( −0.10)3 ( −0.10) 2 ⎞⎤

= 5 ⎢⎜⎜ + ⎟⎟ − ⎜⎜ + ⎟⎟⎥ =
⎣ ⎝ 3 2 ⎠ ⎝ 3 2 ⎠⎦

⎡⎛ 0.103 0.102 ⎞ ⎛ 0.103 0.102 ⎞⎤

= 5 ⎢⎜⎜ + ⎟ − ⎜− + ⎟⎥ =
⎣⎝ 3 2 ⎟⎠ ⎜⎝ 3 2 ⎟⎠⎦

⎛ 0.103 0.10 2 0.103 0.10 2 ⎞ ⎛ 0.103 ⎞ ⎛2⎞

= 5 ⎜⎜ + + − ⎟⎟ = 5 ⎜⎜ 2 ⋅ ⎟⎟ = 5 ⎜ ⎟ (0.001) =
⎝ 3 2 3 2 ⎠ ⎝ 3 ⎠ ⎝ 3⎠

⎛ 2 ⎞⎛ 1 ⎞ ⎛ 1 ⎞⎛ 1 ⎞ 1
= 5⎜ ⎟ ⎜ ⎟ = ⎜ ⎟⎜ ⎟= = 0.00333
⎝ 3 ⎠ ⎝ 1000 ⎠ ⎝ 3 ⎠ ⎝ 100 ⎠ 300

ANSWER: C

E (Y 2 ) =
0.10
23. ∫ y 2 f ( y ) dy =
y = −0.10

0.10
⎡ y 4 y3 ⎤
= ∫ y ⋅ 5 ( y + 1) dy = 5 ∫ (y + y )dy = 5 ⎢ + ⎥
0.10 0.10
2 3 2
=
y = −0.10 y = −0.10 ⎣4 3 ⎦ − 0.10

⎡⎛ 0.104 0.103 ⎞ ⎛ ( −0.10) 4 ( −0.10)3 ⎞⎤

= 5 ⎢⎜⎜ + ⎟⎟ − ⎜⎜ + ⎟⎟⎥ =
⎣ ⎝ 4 3 ⎠ ⎝ 4 3 ⎠⎦

⎡⎛ 0.104 0.103 ⎞ ⎛ 0.104 0.103 ⎞⎤

= 5 ⎢⎜⎜ + ⎟−⎜ − ⎟⎥ =
⎣⎝ 4 3 ⎟⎠ ⎜⎝ 4 3 ⎟⎠⎦

13
 ⎛ 0.10 4 0.103 0.10 4 0.103 ⎞ ⎛ 0.103 ⎞ ⎛2⎞
= 5 ⎜⎜ + − + ⎟⎟ = 5 ⎜⎜ 2 ⋅ ⎟⎟ = 5 ⎜ ⎟ (0.001) =
⎝ 4 3 4 3 ⎠ ⎝ 3 ⎠ ⎝ 3⎠

⎛ 2 ⎞⎛ 1 ⎞ ⎛ 1 ⎞⎛ 1 ⎞ 1
= 5⎜ ⎟ ⎜ ⎟ = ⎜ ⎟⎜ ⎟= = 0.00333
⎝ 3 ⎠ ⎝ 1000 ⎠ ⎝ 3 ⎠ ⎝ 100 ⎠ 300

ANSWER: E

var (Y ) = E (Y 2 ) − [E (Y )] =
2
24.

= 0.00333 − 0.003332 = 0.00333 − 0.00001 = 0.00332

ANSWER: A

25. If X and Y are continuous random variables and f ( x, y ) is their joint probability
distribution, the general definition of expected value of the product of X and Y is

b d
E ( XY ) = ∫ ∫ xy f (x, y ) dy dx
x = a y =c

f ( x, y ) = (x y + x + y + 1), − 0.05 < x < 0.15 , − 0.10 < y < 0.10 .

500
In our case,
21
So we have

0.15 0.10
E ( XY ) = (x y + x + y + 1)dy dx =
500
∫ ∫
x = −0.05 y = −0.10
xy ⋅
21

(x y + x 2 y + xy 2 + xy )dy dx
0.15 0.10
500 2 2
= ∫ ∫
x = −0.05 y = −0.10
21

ANSWER: B

26. Let us first check for the possible independence of the random variables X and Y .
To check for independence of X and Y , it is necessary to verify whether
f ( x ) f ( y ) = f ( x, y ) for any values x and y within their ranges of definition:

f ( x ) f ( y ) = f (x, y ) ∀ x, y within their ranges of definition

Since we know from answers 16 and 21 that

14
 f (x ) = (x + 1) , f ( y ) = 5 ( y + 1) and f (x, y ) = 500 (x y + x + y + 1)
100
21 21

we have

⎡100
f (x ) f ( y ) = ⎢ (x + 1)⎤⎥ ⋅ [5 ( y + 1)] = 500 (x + 1)( y + 1) =
⎣ 21 ⎦ 21

=
500
(x y + x + y + 1) = f (x, y ), ∀ − 0.05 < x < 0.15, − 0.10 < y < 0.10
21

Therefore, X and Y are independent.

As a consequence of the independence of X and Y

E ( XY ) = E ( X )E (Y )

We know from answers 17 and 22 that E ( X ) = 0.05317 and E (Y ) = 0.00333 .

Therefore

E ( XY ) = 0.05317 ⋅ 0.00333 = 0.00018

ANSWER: D

27. As a consequence of the independence of X and Y

E ( XY ) = E ( X )E (Y ) ⇔ E ( XY ) − E ( X )E (Y ) = 0

Therefore

cov ( X , Y ) = E ( XY ) − E ( X )E (Y ) = 0

ANSWER: B

28. As a consequence of the independence of X and Y

cov ( X , Y ) = 0

Therefore

cov( X , Y ) 0
ρX ,Y = = =0
SD ( X ) SD (Y ) SD ( X ) SD (Y )

15
 ANSWER: D

29. We can apply the formula for the expected value of the sum of two functions of X
and Y :

E [g ( X ) + h(Y ) ] = E [g ( X ) ] + E [h(Y ) ]

In our case, g ( X ) = 0.70 X and h(Y ) = 0.30Y .

We know from answers 17 and 22 that E ( X ) = 0.05317 and E (Y ) = 0.00333 . So

for the expected value of the rate of return of the portfolio R we have:

E (R ) = E (0.70 X + 0.30Y ) =

= E (0.70 X ) + E (0.30Y ) =

= 0.70 E ( X ) + 0.30 E (Y ) =

= 0.70 (0.05317 ) + 0.30 (0.00333) =

= 0.03722 + 0.00100 = 0.03822

ANSWER: E

30. If (a + bX ) is a linear combination of X , (c + dY ) is a linear combination of Y , and

R is their sum, i.e., R = (a + bX ) + (c + dY ) – where a, b, c and d are constants – the
variance of R , var (R ) , is given by the formula for the variance of linear combinations
of two random variables:

var (R ) = b2 var ( X ) + d 2 var (Y ) + 2 b d cov( X , Y )

Since R = 0.70 X + 0.30Y , in our case a = 0 , b = 0.70 , c = 0 and d = 0.30 . We

know from answers 19, 24 and 27 that var ( X ) = 0.00332 , var (Y ) = 0.00332 and
cov ( X , Y ) = 0 . So we have:

var (R ) = 0.702 var ( X ) + 0.302 var (Y ) + 2 (0.70 )(0.30 )cov( X ,Y ) =

= 0.49 (0.00332 ) + 0.09 (0.00332 ) + 0 =

= 0.00163 + 0.00030 = 0.00193

ANSWER: C

16
31. For the sample variance of overtime hours per week:

X X−X ( X − X )2
57 6.0625 36.7539
56 5.0625 25.6289
52 1.0625 1.1289
44 -6.9375 48.1289
46 -4.9375 24.3789
53 2.0625 4.2539
44 -6.9375 48.1289
44 -6.9375 48.1289
48 -2.9375 8.6289
51 0.0625 0.0039
55 4.0625 16.5039
48 -2.9375 8.6289
63 12.0625 145.5039
53 2.0625 4.2539
51 0.0625 0.0039
50 -0.9375 0.8789
∑ = 815 ∑ = 420.9375

n 16

∑ Xi ∑X i
815
X = i =1
= i =1
= = 50.9375
n 16 16

∑ (X − X)
n 16

∑ (X − 50.9375)
2 2
i i
420.9375
s2 = i =1
= i =1
= = 28.0625
n −1 16 − 1 15

ANSWER: C

32. It is known that the random variable X (overtime hours per week) is normally
distributed and that the sample is small ( n = 16 ). The object of our inference is the
population variance σ 2 .

Therefore the statistic used to obtain a confidence interval for the population variance
is given by

(n − 1) s 2
σ2

ANSWER: D

17
33. It is known that the random variable X (overtime hours per week) is normally
distributed and that the sample is small ( n = 16 ).

Therefore the above statistic follows the χ 2 distribution:

(n − 1) s 2 = χ 2 ~ χ 2 distribution with ν = n − 1 = 16 − 1 = 15 degrees of freedom

σ 2

ANSWER: A

34. The 95% confidence interval for the population variance σ 2 is given by

⎛ (n − 1) s 2 (n − 1) s 2 ⎞
σ =⎜ 2
2
, ⎟
⎜ χ 0.025, 15 χ 02.975, 15 ⎟
⎝ ⎠

ANSWER: B

35. The statistic used to obtain the confidence interval follows the χ 2 distribution:

χ 2 ~ χ 2 distribution with ν = 15 degrees of freedom

The required confidence level for the interval is 1 − α = 95% .

Therefore, the critical values for the confidence interval are:

χ 02.975, 15 = 6.262 , χ 02.025, 15 = 27.488

2.5% 95% 2.5%

0 χ 152
6 .262 27.488

ANSWER: A

18
 ⎛ (n − 1) s 2 (n − 1) s 2 ⎞ ⎛ (16 − 1) 28.0625 (16 − 1) 28.0625 ⎞
36. σ2 =⎜ , ⎟=⎜ , ⎟=
⎜ χ 02.025, 15 χ 02.975, 15 ⎟ ⎝ 27 .488 6 .262 ⎠
⎝ ⎠

⎛ 420.9375 420.9375 ⎞
=⎜ , ⎟ = (15.3135, 67.2209)
⎝ 27.488 6.262 ⎠

ANSWER: A

37. The variance of overtime hours per week in the population is some value between
15.3135 and 67.2209 with a 95% probability. Since this interval includes the value
25 , the confidence interval obtained includes the possibility that the population
variance σ 2 is 25 , as expected. There is therefore no evidence (with 95%
confidence) that the variance has changed.

ANSWER: D

38. It is known that, for method 1, the sample variance is s12 = 19.0286 and the sample
size n1 = 15 ; for method 2, the sample variance is s 22 = 24.9318 and the sample size
n 2 = 12 . So the value of the pooled estimate of the common population variance is:

s 2p =
(n1 − 1) s12 + (n2 − 1) s 22 =
(15 − 1)19.0286 + (12 − 1) 24.9318 =
n1 + n2 − 2 15 + 12 − 2

=
(14)19.0286 + (11) 24.9318 = 21.626
25

ANSWER: B

39. For the test of the null hypothesis that the population mean test score with method 2 is
the same as with method 1, against the alternative that the population mean test score
with method 2 is worse than with method 1, the null and alternative hypotheses are:

H 0 : µ 2 = µ1 H 1 : µ 2 < µ1

µ1 − µ 2 = 0 µ1 − µ 2 > 0

ANSWER: C

19
40. It is known that the two populations have a common population variance, i.e.,
σ 12 = σ 22 = σ 2 . For two small samples ( n1 = 15 and n2 = 12 ), sampling theory
establishes that the sampling distribution of the difference between the two sample
means ( X 1 − X 2 ) is:

⎛ σ2 σ2 ⎞
(X − X ) ~ N ⎜ µ
⎜ 1 − µ , + ⎟
n1 n2 ⎟⎠
1 2 2
⎝

The common population variance σ 12 = σ 22 = σ 2 is unknown and we replace it with its

sample estimate, the pooled estimate of the common population variance s 2p .

Therefore the test statistic is given by

(X 1 − X 2 ) − (µ1 − µ 2 )0
=
(X 1 − X 2 ) − (µ1 − µ 2 )0
s 2p s 2p 1 1
+ sp +
n1 n2 n1 n2

ANSWER: B

41. It is known that, for method 1, the sample mean is X 1 = 47.80 and the sample size
n1 = 15 ; for method 2, the sample mean is X 2 = 46.75 and the sample size n 2 = 12 .
From answer 38 we know that the pooled estimate of the common population
variance is s 2p = 21.626 . So the value of the test statistic is:

(X 1 − X 2 ) − (µ1 − µ 2 )0
=
(47.80 − 46.75) − 0 = 1.05
=
s 2p s 2p 21.626 21.626 1.4417 + 1.8022
+ +
n1 n2 15 12

1.05 1.05
= = = 0.5830
3.2439 1.8011

ANSWER: D

42. The common population variance σ 12 = σ 22 = σ 2 is unknown and we replace it with

the pooled estimate s 2p . The two samples are small ( n1 = 15 and n 2 = 12 ).

Therefore the test statistic follows the t distribution:

20
 (X 1 − X 2 ) − (µ1 − µ 2 )0
= t ~ t distribution with ν = n1 + n2 − 2 = 15 + 12 − 2 = 25
1 1
sp + degrees of freedom
n1 n2

ANSWER: C

43. The test statistic follows the t distribution:

t ~ t distribution with ν = 25 degrees of freedom

Since the alternative hypothesis takes the form H 1 : µ1 − µ 2 > 0 , the test is one-
tailed.

The required significance level for the test is α = 1% .

Therefore, the critical value for the test is:

t 0.01, 25 = 2.4851

1%
99%

0 t 25
2.4851

ANSWER: E

44.

1%
99%

0 t 25
2.4851

Do not Reject
reject H 0 H0

21
 • if t ≥ 2.4851 , reject H 0
• if t < 2.4851 , do not reject H 0

ANSWER: D

45. Since the test statistic t = 0.5830 < 2.4851 , we do not reject H 0 : µ 2 = µ1 at the 1%
significance level in favour of the alternative H 1 : µ 2 < µ1 . There is no significant
evidence (at the 1% level) that the population mean test score with method 2 is worse
than with method 1.

ANSWER: A

46. If X 1 = 312 is the number of CEOs who believe fear of getting caught or losing one’s
job has a strong influence on ethical behaviour, X 2 = 442 the number of consumers
who believe fear of getting caught or losing one’s job has a strong influence on ethical
behaviour, n1 = 650 the size of the sample of CEOs, and n 2 = 850 the size of the
sample of consumers, sampling theory establishes that the sample proportions
X X
P1 = 1 and P2 = 2 are unbiased point estimators of the corresponding population
n1 n2
proportions π 1 and π 2 , i.e., E (P1 ) = π 1 and E (P2 ) = π 2 . Therefore:

X 1 312 X 2 442
P1 = = = 0.48 and P2 = = = 0.52
n1 650 n2 850

are unbiased point estimators of π 1 and π 2 .

ANSWER: E

47. For the test of the null hypothesis that the population proportion of CEOs who believe
fear of getting caught or losing one’s job has a strong influence on ethical behaviour
is equal to 50% , against the alternative that the population proportion of CEOs with
such belief is smaller than 50% – i. e., that only a minority of CEOs believe fear of
getting caught or losing one’s job has a strong influence on ethical behaviour – the
null and alternative hypotheses are:

H 0 : π 1 = 0.50 H 1 : π 1 < 0.50

ANSWER: C

22
48. For the large sample of CEOs ( n1 = 650 ) taken from a finite population ( N1 = 1035 )
without replacement, the sampling ratio is:

n1 650
= = 0.63 > 0.10 (10%)
N 1 1035

⎛ N − n1 ⎞
Therefore we need to use the finite population correction of the variance: ⎜⎜ 1 ⎟⎟ .
⎝ N1 − 1 ⎠

Sampling theory establishes that the sampling distribution of the sample proportion P1
is:
⎛ π (1 − π 1 ) ⎛ N 1 − n1 ⎞ ⎞⎟
P1 ~ N ⎜⎜ π 1 , 1 ⎜⎜ ⎟⎟
⎟
a
⎝ n1 ⎝ N 1 − 1 ⎠⎠

The population proportion π 1 is unknown. So we use the hypothetical value of π 1

under the null hypothesis, π 1 0 , in estimating the standard error of the sample
proportion P1 , SE (P1 ) .

Therefore the test statistic is given by

P1 − π 1 0 P1 − π 1 0
=
SE ( P1 ) π 1 0 (1 − π 1 0 ) ⎛ N 1 − n1 ⎞
⎜⎜ ⎟⎟
n1 ⎝ N1 − 1 ⎠

ANSWER: B

49. It is known that the sample size is n1 = 650 , the finite population size N1 = 1035 , and
the hypothetical value of π 1 under the null hypothesis π 1 0 = 0.50 . It is also known
from answer 46 that the sample proportion of CEOs who believe fear of getting
caught or losing one’s job has a strong influence on ethical behaviour is P1 = 0.48 . So
the value of the test statistic is:

P1 − π 1 0 0.48 − 0.50
= =
π 1 0 (1 − π 1 0 ) ⎛ N 1 − n1 ⎞ 0.50 (1 − 0.50) ⎛ 1035 − 650 ⎞
⎜⎜ ⎟⎟ ⎜ ⎟
n1 ⎝ N1 − 1 ⎠ 650 ⎝ 1035 − 1 ⎠

− 0.02 − 0.02 − 0.02

= = = =
0.25 ⎛ 385 ⎞ 0.05 ⎛ 7 ⎞ 0.35
⎜ ⎟ ⎜ ⎟
650 ⎝ 1034 ⎠ 26 ⎝ 94 ⎠ 2444

23
 − 0.02 − 0.02
= = = −1.67
0.000143 0.0120

ANSWER: E

50. Since the sampling distribution of the sample proportion P1 is normal:

⎛ π (1 − π 1 ) ⎛ N 1 − n1 ⎞ ⎞⎟
P1 ~ N ⎜⎜ π 1 , 1 ⎜⎜ ⎟⎟
⎟
a
⎝ n1 ⎝ N1 − 1 ⎠ ⎠

the test statistic follows the standard normal distribution:

P1 − π 1 0
= Z ~ N (0, 1)
π 1 0 (1 − π 1 0 ) ⎛ N 1 − n1 ⎞
⎜⎜ ⎟⎟
n1 ⎝ N1 − 1 ⎠

ANSWER: B

51. The test statistic follows the standard normal distribution:

Z ~ N (0, 1)

Since the alternative hypothesis takes the form H 1 : π 1 < 0.50 , the test is one-tailed
(left-hand side tail).

The required significance level is α = 5% .

Therefore, the critical value for the test is:

z 0.95 = −1.6449

5% 1

95%

Z
− 1.6449 0

ANSWER: A

24
52.

5% 1

95%

Z
− 1.6449 0

Reject Do not
H0 reject H 0

• if Z ≤ −1.6449 , reject H 0
• if Z > −1.6449 , do not reject H 0

ANSWER: B

53. Since the test statistic Z = −1.67 < −1.6449 , we reject H 0 : π 1 = 0.50 at the 5%
significance level in favour of the alternative H 1 : π 1 < 0.50 . There is significant
evidence (at the 5% level) that the population proportion of CEOs who believe fear
of getting caught or losing one’s job has a strong influence on ethical behaviour is
smaller than 50% – i. e., that only a minority of CEOs hold such belief.

ANSWER: C

54. For the test of the null hypothesis that the same population proportion of consumers
( π 2 ) and CEOs ( π 1 ) believe fear of getting caught or losing one’s job has a strong
influence on ethical behaviour, against the alternative hypothesis that a higher
population proportion of consumers than CEOs believe fear of getting caught or
losing one’s job has a strong influence on ethical behaviour, the null and alternative
hypotheses are:

H 0 : π 2 = π1 H1 : π 2 > π 1

that is

H 0 : π1 − π 2 = 0 H1 : π 1 − π 2 < 0

ANSWER: C

25
55. For the large sample of CEOs ( n1 = 650 ) taken from a finite population ( N1 = 1035 )
without replacement, the sampling ratio is:

n1 650
= = 0.63 > 0.10 (10%)
N 1 1035

⎛ N − n1 ⎞
Therefore we need to use the finite population correction of the variance: ⎜⎜ 1 ⎟⎟ .
⎝ N1 − 1 ⎠

For the large sample of consumers ( n 2 = 850 ) taken from a very large population
without replacement, we do not need to use the finite population correction of the
variance.

Sampling theory establishes that the sampling distribution of the difference between
the two sample proportions (P1 − P2 ) is:

⎛ ⎞
(P1 − P2 ) ~a N ⎜⎜ π 1 − π 2 , π 1 (1 − π 1 ) ⎜⎜ N1 − n1 ⎟⎟ + π 2 (1 − π 2 ) ⎟⎟
⎛ ⎞
⎝ n1 ⎝ N1 − 1 ⎠ n2 ⎠

The population proportions π 1 and π 2 are unknown. So we use the sample

proportions P1 and P2 , the unbiased point estimators of π 1 and π 2 , in estimating the
standard error of the difference between the two sample proportions (P1 − P2 ) ,
SE (P1 − P2 ) .

Therefore the test statistic is given by

(P1 − P2 ) − (π 1 − π 2 )0 (P1 − P2 ) − (π 1 − π 2 )0
=
SE (P1 − P2 ) P1 (1 − P1 ) ⎛ N 1 − n1 ⎞ P2 (1 − P2 )
⎜
⎜ N −1 ⎟ ⎟+
n1 ⎝ 1 ⎠ n2

ANSWER: C

56. It is known that, for CEOs, the sample size is n1 = 650 , the finite population size is
N 1 = 1035 and, from answer 46, the sample proportion who believes fear of getting
caught or losing one’s job has a strong influence on ethical behaviour is P1 = 0.48 ; for
consumers, the sample size is n 2 = 850 and, from answer 46, the sample proportion
who believes fear of getting caught or losing one’s job has a strong influence on
ethical behaviour is P2 = 0.52 . It is also known that the difference between
population proportions under the null hypothesis is (π 1 − π 2 )0 = 0 . So the value of the
test statistic is:

26
 (P1 − P2 ) − (π 1 − π 2 )0 (0.48 − 0.52) − 0
= =
P1 (1 − P1 ) ⎛ N 1 − n1 ⎞ P2 (1 − P2 ) 0.48 (1 − 0.48) ⎛ 1035 − 650 ⎞ 0.52 (1 − 0.52)
⎜ ⎟+ ⎜ +
⎟
⎜ N −1 ⎟ 650 ⎝ 1035 − 1 ⎠ 850
n1 ⎝ 1 ⎠ n2

− 0.04
= =
0.2496 ⎛ 385 ⎞ 0.2496
⎜ ⎟+
650 ⎝ 1034 ⎠ 850

− 0.04 − 0.04
= = =
0.0048 ⎛ 7 ⎞ 0.1248 0.0336 0.1248
⎜ ⎟+ +
5 ⎝ 47 ⎠ 425 235 425

− 0.04 − 0.04
= = =
0.000143 + 0.000294 0.000437

− 0.04
= = −1.91
0.0209

ANSWER: B

57. Since the sampling distribution of the difference between the two sample proportions
(P1 − P2 ) is normal:

⎛ ⎞
(P1 − P2 ) ~a N ⎜⎜ π 1 − π 2 , π 1 (1 − π 1 ) ⎜⎜ N1 − n1 ⎟⎟ + π 2 (1 − π 2 ) ⎟⎟
⎛ ⎞
⎝ n1 ⎝ N1 − 1 ⎠ n2 ⎠

the test statistic follows the standard normal distribution:

(P1 − P2 ) − (π 1 − π 2 )0
= Z ~ N (0, 1)
P1 (1 − P1 ) ⎛ N 1 − n1 ⎞ P2 (1 − P2 )
⎜⎜ N −1 ⎟ ⎟+
n1 ⎝ 1 ⎠ n2

ANSWER: A

58. The test statistic follows the standard normal distribution:

Z ~ N (0, 1)

Since the alternative hypothesis takes the form H 1 : π 1 − π 2 < 0 , the test is one-
tailed (left-hand side tail).

27
 The required significance level is α = 1% .

Therefore, the critical value for the test is:

z 0.99 = −2.3263

1
1%
99%

0 Z
− 2.3263

ANSWER: D

59.

1
1%
99%

0 Z
− 2.3263

Reject Do not
H0 reject H 0

• if Z ≤ −2.3263 , reject H 0
• if Z > −2.3263 , do not reject H 0

ANSWER: C

60. Since the test statistic Z = −1.91 > −2.3263 , we do not reject H 0 : π 2 = π 1 at the
1% significance level in favour of the alternative H 1 : π 2 > π 1 . There is no
significant evidence (at the 1% level) that a higher population proportion of
consumers ( π 2 ) than CEOs ( π 1 ) believe fear of getting caught or losing one’s job has
a strong influence on ethical behaviour.

ANSWER: A

28