Statistics 430

HW #6 Solutions
Hyunseung Kang, Emil Pitkin, and Yun Zhang
October 31, 2011

Ch 5 Problem 3

Neither is a density function. Both 2x x3 and 2x x2 attain positive and negative values
in the interval 0 < x < 52 , so no matter the value of C, f (x) is negative somewhere in the
interval. In contrast, density functions are, by definition, non-negative.

Ch 5 Problem 5

When the tanks capacity is the 99th percentile of sales volume, the probability of the supplys
being exhausted in a given week is .01. Let C denote the tanks capacity. We wish for C to
satisfy
Z C
(1 x)4 dx = .99
5
0
C
5
(1 x) 
= .99
0

(1 C)5 = .01

C =1

.01 = .602

The tanks capacity should therefore be 602 gallons.

Ch 5 Problem 7

We will solve two simultaneous equations. We are equipped with two pieces of knowledge:
1) E[X] = 53 and
2) the density must integrate to 1.
Expectation equation:
Z

3
x(a + bx2 )dx =
5
0

1
b 
3
a
( x 2 + x4 )  =
2
4
5
0
a b
3
+
=
2 4
5
1

Density equation:
Z

a + bx2 dx = 1
0
b 3 1
ax + x  = 1
3 0
b
a+
= 1
3

Solving the two equations for a and b yields a =

3
5

and b =

6
5

Ch 5 Problem 8*

Integrate by parts:
Z

x2 ex dx

E[X] =
0

= x

1
e
1
=
e
1
=
e

Z 1
1
e  2
xex dx
0
0


1 Z 1
x 
x
2 xe 
e dx
0
0


1
1
x 
2 +e 
e
0



1
1
2 +
1
e
e

2 x 

2 hours

Ch 5 Problem 11

Let X be the location of the chosen point. Then X U nif [0, L]. We can assume without
loss of generality that L equals 1.




X
1
X
P
<
+P
>4
= P (4X < 1 X) + P (X > 4 4X)
1X
4
1X

 

1
4
+ X>
= P X<
5
5
= .2 + .2 = .4

Ch 5 Problem 12

The more efficient placement would minimize the expected distance between breakdown point
and bus service station. Let B denote the breakdown point. Under the present arrangement,
the minimum distance M between B and a service station is given by

M =

E[M ] =

=
=

B
0 < B < 25

50 B
25 < B < 50
B

50
50 < B < 75

100 B
75 < B < 100
Z 50
Z 25
1
1
(50 B)
B
dB +
dB
100
100
25
0
Z 75
Z 100
1
1
(B 50)
(100 B)
+
dB +
dB
100
100
50
75
 2


 2


 
1
B 25
B 2 50
B
B 2 100
75
50B  + 100B
 +

 + 50B
100 2 0
2 25
2
2 75
50



502 252
1 252 02
+ 50 25
100
2
2
 2
 

2
1
75 50
1002 752
+
50 25 + 50 25
100
2
2
12.5 miles

We could have avoided the mess and arrived at the same happy ending by arguing by symmetry.
Under the alternate arrangement, the minimum distance M between B and a service station
is given by

B
0 < B < 12.5

25 B
12.5 < B < 25

B 25
25 < B < 37.5

50 B
37.5 < B < 50
M =
B 50
50 < B < 62.5

75 B
62.5 < B < 75

75
75 < B < 87.5

100 B
37.5 < B < 50
With analagous calculations and reasoning, we find that the expected distance between the
breakdown point and nearest rest stop is 6.25 miles . Indeed, the alternate arrangement is
more efficient.

Ch 5 Problem 13

Let A be the number of minutes past 10 oclock when the bus arrives. A U nif [0, 30].
Recall that the CDF of a U nif [a, b] random variable is given by F (x) = xa
.
ba

a
P (A > 10) = 1 F (10)
= 1

2
10 0
=
30 0
3

b
P (A > 25 A > 10)
P (A > 15)
P (A > 25)
=
P (A > 15)
1
= 6/ 21

P (A > 25|A > 15) =

1
3

Ch5 Problem 14*

If X U (0, 1), then f (x) = 1 for 0 < x < 1 and F (x) = x for 0 < x < 1. Since 0 < X < 1,
so 0 < X n < 1. By using Proposition 2.1,
Z 1
n
P (X n > t)dt
E(X ) =
Z0 1
P (X > t1/n )dt
=
Z0 1
1 P (X t1/n )dt
=
Z0 1
1 F (t1/n )dt
=
Z0 1
1 t1/n dt
=
0
"
#1
1
+1
n
t
= t 1
+1
n
0
n
= 1
n+1
1
=
.
n+1
By using the definition of expectation, we first have to derive the pdf for f (xn ). To do
we find the cumulative distribution function and take the derivative (aka FX n (x) =
Rthis,
x
d
f (x)dx dx
FX n (x) = f (x), by the Fundamental Theorem of Calculus)
0
FX n (x) = P (X n < x) = P (X < x1/n ) = x1/n

1 1/n1
x
n

Then, our expectation is:

Z

E(X ) =
=
=
=
=

1
x x1/n1 dx
n
Z0 1
1 1
x n dx
0 n
" 1 #1
1 x n +1
n n1 + 1
0

1
1
1
n n +1
1
.
n+1

Ch 5 Problem 15

a


X
5 10
P (X > 5) = P
>

6
= P (Z > .83)
= .798

b
16 10
4 10
) P (Z <
)
6
6
= P (Z < 1) P (Z < 1)
= .683

P (4 < X < 16) = P (Z <

c
8 10
)
6
= P (Z < .33)
= .369

P (X < 8) = P (Z <

d
20 10
)
6
= P (Z < 1.67)
= .952

P (X < 20) = P (Z <

e
16 10
)
6
= P (Z > 1)
= .159

P (X > 16) = P (Z >

10

Ch 5 Problem 17

Let D denote the distance from the shot to the target. We are given that D U nif [0, 10].
The winnings W are distributed as follows:

10
0<D<1

5
1<D<3
W =
3
3<D<5

0
5 < D < 10
X
wi p(wi )
E[W ] =
wi

= 10
=

11

2
2
1
1
+5
+3 +0
10
10
5
2

2.6

Ch5 Problem 18*

Let X be a normal random variable with mean 5 and variance 2 . X N (5, 2 ). If

0.2 = P (X > 9), then
0.8 = P (X 9)


95
X 5
X 5

= Z N (0, 1)
, where
= P

 
4
=

From a standard normal table, we have 1 (0.8) = 0.842. Thus

4
= 0.842 = 4.75.

Hence V ar(X) = 2 = 22.56.

12

Ch5 Theoretical Problem 2*

In order to show that

Z

Z
P (Y > y)dy

E[Y ] =
0

P (Y < y)dy,
0

it is equivalent to show that

Z

xfY (x)dx
P (Y < y)dy =

0
Z
Z
P (Y > y)dy =
xfY (x)dx.
0

Now we are showing (1),

Z

Z

LHS =
y=0
Z 0

x=
Z x

=
x=
0

fY (x)dx dy

1dy fY (x)dx

y=0

xfY (x)dx

=
x=

= RHS.

And showing (2),

Z

Z

LHS =
y=0
Z

x=y
Z x

Zx=0

y=0


fY (x)dx dy

1dy fY (x)dx

xfY (x)dx

=
x=0

= RHS.
Combining (1) and (2), we have
Z
E[Y ] =
yfY (y)dy

Z 0
Z
=
xfY (x)dx +

xfY (x)dx, where x, y are dummy

= (2) (1).

13

Ch5 Theoretical Problem 5*

For a nonnegative random variable X, (X n nonnegative) we have

Z
n
E[X ] =
P (X n > t)dt.
t=0

Using change of variables t = xn , {X n > xn } {X > x} and

7

dt
dx

= nxn1 ,

(1)
(2)

E[X ] =

P (X > x)
Zx=0

dt
dx
dx

P (X > x)nxn1 dx, as required.

x=0

14

Ch5 Theoretical Problem 8

15

Ch5 Theoretical Problem 13

We derive all of these medians using the definition of the median (i.e. m is a median of X
if F (m) = 12 ), rather than intuition (because its obvious, otherwise)

15.1

15.2

15.3

16

Ch5 Theoretical Problem 21

Using x

2x change of variables, we get the following

1
( ) =
2

x 1/2

e x

Z
dx =

where the last equality uses

12 y 2

e
0

R
0

1 2
1 e 2 y dy
2

1 2
y
2

1
2

1/2

Z
ydy =
0

1 2

2e 2 y dy =