CHAPTER 5

Section 5-1

5-1. First, f(x,y) ≥ 0. Let R denote the range of (X,Y).

Then, f ( x, y ) = 14 + 18 + 14 + 14 + 18 = 1
R

5-2 a) P(X < 2.5, Y < 3) = f(1.5,2)+f(1,1) = 1/8+1/4=3/8

b) P(X < 2.5) = f (1.5, 2) + f (1.5, 3) +f(1,1)= 1/8 + 1/4+1/4 = 5/8
c) P(Y < 3) = f (1.5, 2)+f(1,1) = 1/8+1/4=3/8
d) P(X > 1.8, Y > 4.7) = f ( 3, 5) = 1/8

5-3. E(X) = 1(1/4)+ 1.5(3/8) + 2.5(1/4) + 3(1/8) = 1.8125

E(Y) = 1(1/4)+2(1/8) + 3(1/4) + 4(1/4) + 5(1/8) = 2.875

5-4 a) marginal distribution of X

x f(x)
1 1/4
1.5 3/8
2.5 1/4
3 1/8
f XY (1.5, y )
b) fY 1.5 ( y ) = and f X (1.5) = 3/8. Then,
f X (1.5)
y fY 1.5 ( y )
2 (1/8)/(3/8)=1/3
3 (1/4)/(3/8)=2/3
f XY ( x,2)
c) f X 2 ( x) = and fY (2) = 1/8. Then,
f Y ( 2)
x f X 2 ( y)
1.5 (1/8)/(1/8)=1

d) E(Y|X=1.5) = 2(1/3)+3(2/3) =2 1/3

e) Since fY|1.5(y)≠fY(y), X and Y are not independent

5-5 Let R denote the range of (X,Y). Because

f ( x, y ) = c(2 + 3 + 4 + 3 + 4 + 5 + 4 + 5 + 6) = 1 , 36c = 1, and c = 1/36
R

5-6. a) P ( X = 1, Y < 4) = f XY (1,1) + f XY (1,2) + f XY (1,3) = 361 ( 2 + 3 + 4) = 1 / 4

b) P(X = 1) is the same as part a. = 1/4
c) P (Y = 2) = f XY (1,2) + f XY ( 2,2) + f XY (3,2) = 361 (3 + 4 + 5) = 1 / 3
1
d) P ( X < 2, Y < 2) = f XY (1,1) = 36
(2) = 1 / 18

5-1
5-7.
E ( X ) = 1[ f XY (1,1) + f XY (1,2) + f XY (1,3)] + 2[ f XY (2,1) + f XY (2,2) + f XY (2,3)]
+ 3[ f XY (3,1) + f XY (3,2) + f XY (3,3)]
= (1 × 369 ) + (2 × 12
36
) + (3 × 1536 ) = 13 / 6 = 2.167
V ( X ) = (1 − 136 ) 2 9
36
+ (2 − 136 ) 2 12
36
+ (3 − 136 ) 2 15
36
= 0.639
E (Y ) = 2.167
V (Y ) = 0.639

5-8 a) marginal distribution of X

x f X ( x) = f XY ( x,1) + f XY ( x,2) + f XY ( x,3)
1 1/4
2 1/3
3 5/12

f XY (1, y )
b) fY X ( y ) =
f X (1)
y f Y X ( y)

1 (2/36)/(1/4)=2/9
2 (3/36)/(1/4)=1/3
3 (4/36)/(1/4)=4/9
f XY ( x,2) 12
c) f X Y ( x) = and f Y ( 2) = f XY (1,2) + f XY ( 2,2) + f XY (3,2) = 36
= 1/ 3
f Y ( 2)
x f X Y ( x)

1 (3/36)/(1/3)=1/4
2 (4/36)/(1/3)=1/3
3 (5/36)/(1/3)=5/12

d) E(Y|X=1) = 1(2/9)+2(1/3)+3(4/9) = 20/9

e) Since fXY(x,y) ≠fX(x)fY(y), X and Y are not independent.

5-9. f ( x, y ) ≥ 0 and f ( x, y ) = 1
R

5-10. a) P ( X < 0.5, Y < 1.5) = f XY ( −1,−2) + f XY ( −0.5,−1) = 18 + 1

4
= 3
8
3
b) P ( X < 0.5) = f XY ( −1,−2) + f XY ( −0.5,−1) = 8
7
c) P (Y < 1.5) = f XY ( −1,−2) + f XY ( −0.5,−1) + f XY (0.5,1) = 8
5
d) P ( X > 0.25, Y < 4.5) = f XY (0.5,1) + f XY (1,2) = 8

5-2
5-11.
E ( X ) = −1( 18 ) − 0.5( 14 ) + 0.5( 12 ) + 1( 18 ) = 1
8

E (Y ) = −2( 18 ) − 1( 14 ) + 1( 12 ) + 2( 18 ) = 1
4

5-12 a) marginal distribution of X

x f X (x)
-1 1/8
-0.5 ¼
0.5 ½
1 1/8

f XY (1, y )
b) fY X ( y ) =
f X (1)
y f Y X ( y)

2 1/8/(1/8)=1
f XY ( x,1)
c) f X Y ( x) =
f Y (1)
x f X Y ( x)

0.5 ½/(1/2)=1

d) E(Y|X=1) = 0.5

e) no, X and Y are not independent

5-13. Because X and Y denote the number of printers in each category,

X ≥ 0, Y ≥ 0 and X + Y = 4

5-14. a) The range of (X,Y) is

3
y 2
1
0 3
1 2
x

Let H = 3, M = 2, and L = 1 denote the events that a bit has high, moderate, and low
distortion, respectively. Then,

5-3
 x,y fxy(x,y)
0,0 0.85738
0,1 0.1083
0,2 0.00456
0,3 0.000064
1,0 0.027075
1,1 0.00228
1,2 0.000048
2,0 0.000285
2,1 0.000012
3,0 0.000001
b)
x fx(x)
0 0.970299
1 0.029835
2 0.000297
3 0.000001

f XY (1, y )
c) fY 1 ( y ) = , fx(1) = 0.29835
f X (1)
y fY|1(x)
0 0.09075
1 0.00764
2 0.000161
E(X)=0(0.970299)+1(0.029403)+2(0.000297)+3*(0.000001)=0.03
(or np=3*.01).
f (1, y )
d) fY 1 ( y ) = XY , fx(1) = 0.029403
f X (1)
y fY|1(x)
0 0.920824
1 0.077543
2 0.001632
e) E(Y|X=1)=0(.920824)+1(0.077543)+2(0.001632)=0.080807
f) No, X and Y are not independent since, for example, fY(0)≠fY|1(0).

5-4
5-15 a) The range of (X,Y) is X ≥ 0, Y ≥ 0 and X + Y ≤ 4 . X is the number of pages with
moderate graphic content and Y is the number of pages with high graphic output out of 4.

x=0 x=1 x=2 x=3 x=4

-05
y=4 5.35x10 0 0 0 0
y=3 0.00184 0.00092 0 0 0
y=2 0.02031 0.02066 0.00499 0 0
y=1 0.08727 0.13542 0.06656 0.01035 0
y=0 0.12436 0.26181 0.19635 0.06212 0.00699

b.)
x=0 x=1 x=2 x=3 x=4
f(x) 0.2338 0.4188 0.2679 0.0725 0.0070

c.)
E(X)=
4
xi f ( xi ) = 0(0.2338) + 1(0.4188) + 2(0.2679) + 3(0.7248) = 4(0.0070) = 1.2
0

f XY (3, y )
d.) f Y 3 ( y ) = , fx(3) = 0.0725
f X (3)
y fY|3(y)
0 0.857
1 0.143
2 0
3 0
4 0

e) E(Y|X=3) = 0(0.857)+1(0.143) = 0.143

f) V(Y|X=3) = 02(0.857)+12(0.143)- 0.1432= 0.123
g) no, X and Y are not independent

5-5
5-16 a) The range of (X,Y) is X ≥ 0, Y ≥ 0 and X + Y ≤ 4 . X is the number of defective
items found with inspection device 1 and Y is the number of defective items found with
inspection device 2.

x=0 x=1 x=2 x=3 x=4

-19 -16 -14 -12 -11
y=0 1.94x10 1.10x10 2.35x10 2.22x10 7.88x10
-16 -13 -11 -9 -7
y=1 2.59x10 1.47x10 3.12x10 2.95x10 1.05x10
-13 -11 -8 -6 -5
y=2 1.29x10 7.31x10 1.56x10 1.47x10 5.22x10
-11 -8 -6 -4
y=3 2.86x10 1.62x10 3.45x10 3.26x10 0.0116
y=4 2.37x10-9 1.35x10-6 2.86x10-4 0.0271 0.961

4 4
f (x, y) = ( 0 . 993 ) x ( 0 . 007 ) 4 − x ( 0 . 997 ) y ( 0 . 003 ) 4 − y
x y

For x=1,2,3,4 and y=1,2,3,4

b.)

x=0 x=1 x=2 x=3 x=4

4
f (x, y) = ( 0 . 993 ) x ( 0 . 007 ) 4 − x for x = 1,2,3,4
x
f(x) 2.40 x 10-9 1.36 x 10-6 2.899 x 10-4 0.0274 0.972

c.)since x is binomial E(X)= n(p) = 4*(0.993)=3.972

f XY (2, y )
d.) f Y |2 ( y ) = = f ( y ) , fx(2) = 0.0725
f X ( 2)
y fY|1(y)=f(y
)
0 8.1 x 10-11
1 1.08 x 10-7
2 5.37 x 10-5
3 0.0119
4 0.988

e) E(Y|X=2) = E(Y)= n(p)= 4(0.997)=3.988

f) V(Y|X=2) = V(Y)=n(p)(1-p)=4(0.997)(0.003)=0.0120
g) yes, X and Y are independent.

5-6
Section 5-2

5-17. a) P( X = 2) = f XYZ (2,1,1) + f XYZ (2,1,2) + f XYZ (2,2,1) + f XYZ (2,2,2) = 0.5
b) P( X = 1, Y = 2) = f XYZ (1,2,1) + f XYZ (1,2,2) = 0.35
c) c) P( Z < 1.5) = f XYZ (1,1,1) + f XYZ (1,2,1) + f XYZ (2,1,1) + f XYZ (2,2,1) = 0.5
d)
P ( X = 1 or Z = 2) = P ( X = 1) + P ( Z = 2) − P ( X = 1, Z = 2) = 0.5 + 0.5 − 0.3 = 0.7

e) E(X) = 1(0.5) + 2(0.5) = 1.5

P ( X = 1, Y = 1) 0.05 + 0.10
5-18 a) P ( X = 1 | Y = 1) = = = 0 .3
P(Y = 1) 0.15 + 0.2 + 0.1 + 0.05
P ( X = 1, Y = 1, Z = 2) 0 .1
b) P ( X = 1, Y = 1 | Z = 2) == = = 0 .2
P ( Z = 2) 0.1 + 0.2 + 0.15 + 0.05
P( X = 1, Y = 1, Z = 2) 0.10
c) P( X = 1 | Y = 1, Z = 2) = = = 0 .4
P(Y = 1, Z = 2) 0.10 + 0.15

f XYZ ( x,1,2)
5-19. f X YZ ( x) = and f YZ (1,2) = f XYZ (1,1,2) + f XYZ (2,1,2) = 0.25
f YZ (1,2)

x f X YZ (x)
1 0.10/0.25=0.4
2 0.15/0.25=0.6

5-20 a.) percentage of slabs classified as high = p1 = 0.05

percentage of slabs classified as medium = p2 = 0.85
percentage of slabs classified as low = p3 = 0.10

b.) X is the number of voids independently classified as high X ≥ 0

Y is the number of voids independently classified as medium Y ≥ 0
Z is the number of with a low number of voids and Z ≥ 0
And X+Y+Z = 20

c.) p1 is the percentage of slabs classified as high.

d) E(X)=np1 = 20(0.05) = 1
V(X)=np1 (1-p1)= 20(0.05)(0.95) = 0.95

5-7
5-21. a) P ( X = 1, Y = 17, Z = 3) = 0 Because the point (1,17,3) ≠ 20 is not in the range of
(X,Y,Z).

b)
P ( X ≤ 1, Y = 17, Z = 3) = P ( X = 0, Y = 17, Z = 3) + P ( X = 1, Y = 17, Z = 3)
20!
= 0.05 0 0.8517 0.10 3 + 0 = 0.07195
0!17!3!
Because the point (1,17,3) ≠ 20 is not in the range of (X,Y,Z).
( )
c) Because X is binomial, P ( X ≤ 1) = 020 0.05 0 0.95 20 + ( )0.05 0.95
20
1
1 19
= 0.7358
d.) Because X is binomial E(X) = np = 20(0.05) = 1

5-22 a) The probability is 0 since x+y+z>20

P ( X = 2, Z = 3, Y = 17)
P ( X = 2, Z = 3 | Y = 17) = .
P (Y = 17)
Because Y is binomial, P (Y = 17) = ( )0.85
20
17
17
0.15 3 = 0.2428 .
20! 0.05 2 0.8517 0.10 3
Then, P ( X = 2, Z = 3, Y = 17) = =0
2!3!17! 0.2054
P ( X = 2, Y = 17)
b) P ( X = 2 | Y = 17) = . Now, because x+y+z = 20,
P (Y = 17)
20!
P(X=2, Y=17) = P(X=2, Y=17, Z=1) = 0.05 2 0.8517 0.10 1 = 0.0540
2!17!1!
P( X = 2, Y = 17) 0.0540
P( X = 2 | Y = 17) = = = 0.2224
P(Y = 17) 0.2428
c)
P ( X = 0, Y = 17) P ( X = 1, Y = 17)
E ( X | Y = 17) = 0 +1
P (Y = 17) P (Y = 17)
P ( X = 2, Y = 17) P ( X = 3, Y = 17)
+2 +3
P(Y = 17) P (Y = 17)
0.07195 0.1079 0.05396 0.008994
E ( X | Y = 17) = 0 +1 +2 +3
0.2428 0.2428 0.2428 0.2428
=1

5-23. a) The range consists of nonnegative integers with x+y+z = 4.

b) Because the samples are selected without replacement, the trials are not independent
and the joint distribution is not multinomial.

5-8
 f XY ( x , 2 )
5-24 P ( X = x | Y = 2) =
fY (2)
P (Y = 2) =
( )( )( ) + ( )( )( ) + ( )( )( ) = 0.1098 + 0.1758 + 0.0440 = 0.3296
4
0
5
2
6
2
4
1
5
2
6
1
4
2
5
2
6
0

( )
15
4 ( ) 15
4 ( ) 15
4

P ( X = 0 and Y = 2) =
( )( )( ) = 0.1098
4
0
5
2
6
2

( ) 15
4

P ( X = 1 and Y = 2) =
( )( )( ) = 0.1758
4
1
5
2
6
1

( ) 15
4

P ( X = 2 and Y = 2) =
( )( )( ) = 0.0440
4
1
5
2
6
1

( ) 15
4

x f XY (x,2)
0 0.1098/0.3296=0.3331
1 0.1758/0.3296=0.5334
2 0.0440/0.3296=0.1335

5-25. P(X=x, Y=y, Z=z) is the number of subsets of size 4 that contain x printers with graphics
enhancements, y printers with extra memory, and z printers with both features divided by
the number of subsets of size 4. From the results on the CD material on counting
techniques, it can be shown that
( )( )( )
4
x
5
y
6
z
P ( X = x, Y = y , Z = z ) = for x+y+z = 4.
( ) 15
4

a) P ( X = 1, Y = 2, Z = 1) =
( )( )( ) = 0.1758
4
1
5
2
6
1

( ) 15
4

b) P ( X = 1, Y = 1) = P ( X = 1, Y = 1, Z = 2) =
( )( )( ) = 0.2198 4
1 1
5 6
2

( ) 15
4
c) The marginal distribution of X is hypergeometric with N = 15, n = 4, K = 4.
Therefore, E(X) = nK/N = 16/15 and V(X) = 4(4/15)(11/15)[11/14] = 0.6146.

5-9
5-26 a)
P ( X = 1, Y = 2 | Z = 1) = P ( X = 1, Y = 2, Z = 1) / P ( Z = 1)
= [( )(( )() )] [( ( )( ) )] = 0.4762
4
1
5
2
15
4
6
1
6
1
15
4
9
3

b)
P ( X = 2 | Y = 2 ) = P ( X = 2 , Y = 2 ) / P (Y = 2 )
[
= ( 2 )((152 )() 0 )
4 5

4
6
] [( ()( ) )] = 0 .1334
5
2
10
2
15
4

c) Because X+Y+Z = 4, if Y = 0 and Z = 3, then X = 1. Because X must equal 1,

f X YZ (1) = 1 .

5-27. a)The probability distribution is multinomial because the result of each trial (a dropped
oven) results in either a major, minor or no defect with probability 0.6, 0.3 and 0.1
respectively. Also, the trials are independent

b.) Let X, Y, and Z denote the number of ovens in the sample of four with major, minor,
and no defects, respectively.
4!
P ( X = 2, Y = 2, Z = 0) = 0.6 2 0.3 2 0.10 = 0.1944
2!2!0!
4!
c. ) P ( X = 0, Y = 0, Z = 4) = 0.6 0 0.3 0 0.14 = 0.0001
0!0!4!

5-28 a.) fXY ( x, y) = fXYZ ( x, y, z) where R is the set of values for z such that x+y+z = 4. That is,
R
R consists of the single value z = 4-x-y and
4!
f XY ( x, y ) = 0.6 x 0.3 y 0.14 − x − y for x + y ≤ 4.
x! y!(4 − x − y )!

b.) E ( X ) = np1 = 4(0.6) = 2.4

c.) E (Y ) = np 2 = 4(0.3) = 1.2

P ( X = 2, Y = 2) 0.1944
5-29 a.) P ( X = 2 | Y = 2) = = = 0.7347
P (Y = 2) 0.2646
4
P (Y = 2) = 0.3 2 0.7 4 = 0.2646 from the binomial marginal distribution of Y
2

b) Not possible, x+y+z=4, the probability is zero.

5-10
 c.) P ( X | Y = 2) = P ( X = 0 | Y = 2), P ( X = 1 | Y = 2), P ( X = 2 | Y = 2)
P ( X = 0, Y = 2) 4!
P ( X = 0 | Y = 2) = = 0.6 0 0.3 2 0.12 0.2646 = 0.0204
P (Y = 2) 0!2!2!
P ( X = 1, Y = 2) 4!
P ( X = 1 | Y = 2) = = 0.610.3 2 0.11 0.2646 = 0.2449
P (Y = 2) 1!2!1!
P ( X = 2, Y = 2) 4!
P ( X = 2 | Y = 2) = = 0.6 2 0.3 2 0.10 0.2646 = 0.7347
P (Y = 2) 2!2!0!
d.) E(X|Y=2)=0(0.0204)+1(0.2449)+2(0.7347) = 1.7143

5-30 Let X,Y, and Z denote the number of bits with high, moderate, and low distortion. Then,
the joint distribution of X, Y, and Z is multinomial with n =3 and
p1 = 0.01, p2 = 0.04, and p3 = 0.95 .
a)
P ( X = 2, Y = 1) = P ( X = 2, Y = 1, Z = 0)
3!
= 0.0120.0410.950 = 1.2 × 10 − 5
2!1!0!
3!
b) P ( X = 0, Y = 0, Z = 3) = 0.0100.0400.953 = 0.8574
0!0!3!

5-31 a., X has a binomial distribution with n = 3 and p = 0.01. Then, E(X) = 3(0.01) = 0.03
and V(X) = 3(0.01)(0.99) = 0.0297.

b. first find P ( X | Y = 2)
P (Y = 2) = P ( X = 1, Y = 2, Z = 0) + P ( X = 0, Y = 2, Z = 1)
3! 3!
= 0.01(0.04) 2 0.95 0 + 0.010 (0.04) 2 0.951 = 0.0046
1!2!0! 0!2!1!
P ( X = 0, Y = 2) 3!
P ( X = 0 | Y = 2) = = 0.010 0.04 2 0.951 0.004608 = 0.98958
P (Y = 2) 0!2!1!

P ( X = 1, Y = 2) 3!
P ( X = 1 | Y = 2) = = 0.0110.04 2 0.95 0 0.004608 = 0.01042
P (Y = 2) 1!2!1!

E ( X | Y = 2) = 0(0.98958) + 1(0.01042) = 0.01042

V ( X | Y = 2) = E ( X 2 ) − ( E ( X )) 2 = 0.01042 − (0.01042) 2 = 0.01031

5-11
5-32 a.) Let X,Y, and Z denote the risk of new competitors as no risk, moderate risk, and very
high risk. Then, the joint distribution of X, Y, and Z is multinomial with n =12 and
p1 = 0.13, p 2 = 0.72, and p 3 = 0.15 . X, Y and Z ≥ 0 and x+y+z=12
b.) P ( X = 1, Y = 3, Z = 1) = 0, not possible since x+y+z≠12

c)
12 12 12
P ( Z ≤ 2) = 0.15 0 0.8512 + 0.151 0.8511 + 0.15 2 0.8510
0 1 2
= 0.1422 + 0.3012 + 0.2924 = 0.7358

5-33 a.) P ( Z = 2 | Y = 1, X = 10) = 0

b.) first get

P( X = 10) = P( X = 10, Y = 2, Z = 0) + P( X = 10, Y = 1, Z = 1) + P( X = 10, Y = 0, Z = 2)
12! 12! 12!
= 0.1310 0.72 2 0.15 0 + 0.1310 0.7210.151 + 0.1310 0.72 0 0.15 2
10!2!0! 10!1!1! 10!0!2!
−8 −8 −9 −8
= 4.72 x10 + 1.97 x10 + 2.04 x10 = 6.89 x10
P ( Z = 0, Y = 2, X = 10) P ( Z = 1, Y = 1, X = 10)
P ( Z ≤ 1 | X = 10) = +
P ( X = 10) P ( X = 10)
12! 12!
= 0.1310 0.72 2 0.15 0 6.89 x10 −8 + 0.1310 0.721 0.151 6.89 x10 −8
10!2!0! 10!1!1!
= 0.9698
c.)
P ( Z = 1, Y = 1, X = 10)
P(Y ≤ 1, Z ≤ 1 | X = 10) =
P( X = 10)
12!
= 0.13100.7210.151 6.89 x10 −8
10!1!1!
= 0.2852

d.
E ( Z | X = 10) = (0(4.72 x10 −8 ) + 1(1.97 x10 −8 ) + 2(2.04 x10 −9 )) / 6.89 x10 −8
= 2.378 x10 −8

5-12
Section 5-3

3 3 3 3 3
2 y2 81
5-34 Determine c such that c xydxdy = c y x2 dy = c(4.5 2
)= 4
c.
0 0 0 0 0

Therefore, c = 4/81.

3 2 3
5-35. a) P ( X < 2, Y < 3) = 4
81
xydxdy = 4
81
(2) ydy = 814 (2)( 92 ) = 0.4444
0 0 0
b) P(X < 2.5) = P(X < 2.5, Y < 3) because the range of Y is from 0 to 3.
3 2.5 3
P( X < 2.5, Y < 3) = 4
81 xydxdy = 4
81 (3.125) ydy = 814 (3.125) 92 = 0.6944
0 0 0
2.5 3 2.5 2.5
y2
c) P (1 < Y < 2.5) = 4
81 xydxdy = 4
81 (4.5) ydy = 18
81 2 =0.5833
1 0 1 1

5-35 d)
2.5 3 2.5
2
P( X > 1.8,1 < Y < 2.5) = 4
81 xydxdy = 4
81 (2.88) ydy = 4
81 (2.88) ( 2.52 −1) =0.3733
1 1.8 1
3 3 3 3
y2
e) E ( X ) = 4
81
x 2 ydxdy = 4
81
9 ydy = 49 2
=2
0 0 0 0

4 0 4
f) P ( X < 0, Y < 4) = 4
81 xydxdy = 0 ydy = 0
0 0 0
3 3
2x
5-36 a) f X ( x) = f XY ( x, y )dy = x 814 ydy = 814 x(4.5) = 9
for 0 < x < 3 .
0 0
4
f XY (1.5, y ) 81
y (1.5)
b) fY 1.5 ( y ) = = 2
= 92 y for 0 < y < 3.
f X (1.5) 9
(1.5)
3 3 3
2 2 2 2y3
c) E(Y|X=1.5) = y y dy = y dy = =6
0
9 90 27 0

2
2 1 2 4 4
d.) P (Y < 2 | X = 1.5) = f Y |1.5 ( y ) = ydy = y 2 = −0 =
0 9 9 0 9 9
4
f XY ( x,2) 81
x ( 2)
e) f X 2 ( x) = = 2
= 92 x for 0 < x < 3.
f Y ( 2) 9
( 2)

5-13
5-37.
3 x+2 3 x+2
y2
c ( x + y )dydx = xy + 2 dx
0 x 0 x

3
= [x( x + 2) + ( x+ 2)2
2 − x2 − x2
2
]dx
0
3
= c (4 x + 2 )dx = 2 x 2 + 2 x [ ] 3

0
= 24c
0
Therefore, c = 1/24.

5-38 a) P(X < 1, Y < 2) equals the integral of f XY ( x, y ) over the following region.

0 1 2 x
0
Then,
1 12 1 1 y2
2 1 3 2
P( X < 1, Y < 2) = ( x + y )dydx = xy + 2
dx = 2 x + 2 − 3 x2 dx =
24 0 x 24 0 x 24 0
1 2 x3
1
x + 2x − 2
= 0.10417
24 0

b) P(1 < X < 2) equals the integral of f XY ( x, y ) over the following region.

0 1 2 x
0
2 x+2 2 x+2
1 1 y2
P (1 < X < 2) = ( x + y )dydx = xy + 2
dx
24 1 x
24 1 x

3
1 1 2
= (4 x + 2)dx = 2 x 2 + 2 x = 16 .
24 0 24 1

5-14
c) P(Y > 1) is the integral of f XY ( x , y ) over the following region.

1 1 1 1
1 1 y2
P (Y > 1) = 1 − P (Y ≤ 1) = 1 − 24
( x + y )dydx = 1 − ( xy + )
0 x 24 0 2 x
1
1
1 1 3 2 1 x2 1 1 3
= 1− x + − x dx = 1 − + − x
24 0 2 2 24 2 2 2 0

= 1 − 0.02083 = 0.9792

d) P(X < 2, Y < 2) is the integral of fXY ( x, y) over the following region.

0 2 x
0

3 x+2 3 x+2
1 1 xy 2
E( X ) = x( x + y )dydx = x2 y + 2 dx
24 0 x
24 0 x

3
1 1 4x3 3 15
= (4 x 2 + 2 x)dx = + x2 =
24 0 24 3 0 8
e)
3 x+2 3 x+2
1 1 xy 2
E( X ) = x( x + y )dydx = x2 y + 2
dx
24 0 x 24 0 x

3
1 1 3 3 15
= (3 x 2 + 2 x)dx = x + x2 =
24 0 24 0 8

5-15
5-39. a) f X ( x) is the integral of f XY ( x, y ) over the interval from x to x+2. That is,
x+2 x+2
1 1 y2 x 1
f X ( x) = ( x + y )dy = xy + 2 = + for 0 < x < 3.
24 x
24 x 6 12
1
f XY (1, y ) 24
(1+ y ) 1+ y
b) f Y 1 ( y ) = f X (1) = 1 1 = for 1 < y < 3.
+
6 12
6
See the following graph,

2 f (y) defined over this line segment

Y|1
1
0 1 2 x
0
3
3 3
1+ y 1 1 y2 y3
c) E(Y|X=1) = y dy = ( y + y 2 )dy = + = 2.111
1 6 61 6 2 3 1
3
3 3
1+ y 1 1 y2
d.) P (Y > 2 | X = 1) = dy = (1 + y )dy = y + =0.5833
2 6 62 6 2 2
f XY ( x , 2 )
e.) f X 2 ( x) = fY ( 2) . Here f Y ( y ) is determined by integrating over x. There are
three regions of integration. For 0 < y ≤ 2 the integration is from 0 to y. For 2 < y ≤ 3
the integration is from y-2 to y. For 3 < y < 5 the integration is from y to 3. Because
the condition is y=2, only the first integration is needed.
y y
1 1 x2 y2
fY ( y) = ( x + y )dx = 2 + xy = 16 for 0 < y ≤ 2 .
24 0 24 0

y
f X|2 (x) defined over this line segment
2
1
0 1 2 x
0
1
( x + 2)
24 x+2
Therefore, fY (2) = 1 / 4 and f X 2 ( x) = = for 0 < x < 2
1/ 4 6

5-16
 3 x 3 3 x
y2 x3 x 4 81
5-40 c xydyd x = c x dx = c dx = c. Therefore, c = 8/81
0 0 0
2 0 0
2 8 8

1 x 1
8 8 x3 8 1 1
5-41. a) P(X<1,Y<2)= xydyd x = dx = = .
81 0 0 81 0 2 81 8 81
2 x 2 2
8 8 x2 8 x4 8 (2 4 − 1) 5
b) P(1<X<2) = xydyd x = x dx = = = .
81 1 0 81 1 2 81 8 1
81 8 27
c)
3
3 x 3 3 3
8 8 x2 −1 8 x x 8 x4 x2
P (Y > 1) = xydyd x = x dx = − dx = −
81 1 1 81 1 2 81 1 2 2 81 8 4 1
4 2 4 2
8 3 3 1 1 1
= − − − = = 0.01235
81 8 4 8 4 81

2 x 2
8 8 x3 8 24 16
d) P(X<2,Y<2) = xydyd x = dx = = .
81 0 0 81 0 2 81 8 81

e.)
3 x 3 x 3 3
8 8 2 8 x2 2 8 x4
E( X ) = x( xy )dyd x = x ydyd x = x dx = dx
81 0 0 81 0 0 81 0 2 81 0 2
8 35 12
= =
81 10 5
f)
3 x 3 x 3
8 8 8 x3
E (Y ) = y ( xy )dyd x = xy 2 dyd x = x dx
81 0 0 81 0 0 81 0 3
3
8 x4 8 35 8
= dx = =
81 0 3 81 15 5

5-17
 x
8 4x 3
5-42 a.) f ( x) = xydy = 0< x<3
81 0 81
8
(1) y
f (1, y ) 81
b.) f Y | x =1 ( y ) = = = 2y 0 < y <1
f (1) 4(1) 3
81
1
1
c.) E (Y | X = 1) = 2 ydy = y 2 =1
0
0
d.) P (Y > 2 | X = 1) = 0 this isn’t possible since the values of y are 0< y < x.
3
8 4y
e.) f ( y ) = xydx = , therefore
81 0 9
8
x ( 2)
f ( x,2) 81 2x
f X |Y = 2 ( x) = = = 0< x<2
f ( 2) 4( 2) 9
9
5-43. Solve for c
∞ x ∞ ∞
c −2x c −2 x
c e − 2 x − 3 y dyd x =
30
(
e 1 − e −3 x d x =
30
)
e − e −5 x d x =
0 0

c 1 1 1
− = c. c = 10
3 2 5 10

5-44 a)
1 x 1 1
− 2 x −3 y 10 − 2 x 10 − 2 x
P( X < 1, Y < 2) = 10 e dyd x = e (1 − e − 3 x )dy = e − e − 5 x dy
0 0
3 0 3 0
1
10 e − 5 x e − 2 x
= − = 0.77893
3 5 2 0

2 x 2
− 2 x −3 y 10 − 2 x
P(1 < X < 2) = 10 e dyd x = e − e −5 x d x
1 0
3 1
b) 2
10 e − 5 x e − 2 x
= − = 0.19057
3 5 2 1
c)
∞ x 1∞
10 − 2 x −9
P(Y > 3) = 10 e − 2 x −3 y dyd x = e (e − e −3 x )dy
3 3
3 3
∞
10 e −5 x e −9 e − 2 x
= − = 3.059 x10 −7
3 5 2 3

5-18
 d)
2
2 x 2
− 2 x −3 y 10 − 2 x 10 e −10 e − 4
P ( X < 2, Y < 2) = 10 e dyd x = e (1 − e −3 x )dx = −
0 0
3 0 3 5 2 0

= 0.9695
∞ x
7
e) E(X) = 10 xe − 2 x − 3 y dyd x =
0 0 10
∞ x
1
f) E(Y) = 10 ye − 2 x − 3 y dyd x =
0 0 5

x
− 2 x −3 y 10e −2 z 10
5-45. a) f ( x ) = 10 e dy = (1 − e − 3 x ) = (e − 2 x − e − 5 x ) for 0 < x
0
3 3
f X ,Y (1, y ) 10e −2 − 3 y
b) f Y \ X =1 ( y ) = = = 3.157e − 3 y 0 < y < 1
f X (1) 10 − 2
(e − e − 5 )
3
1
c) E(Y|X=1 )=3.157 ye -3 y dy=0.2809
0

f X ,Y (x,2 ) 10e −2 x − 6
d) f X |Y = 2 ( x) = = −10
= 2e − 2 x + 4 for 2 < x,
fY ( 2 ) 5e
where f(y) = 5e-5y for 0 < y
∞∞ ∞ ∞
c − 2 x −3 x c 1
5-46 c e − 2 x e −3 y dydx = e (e )dx = e −5 x dx = c c = 15
0 x
30 30 15

5-47. a)
1 2 1
P ( X < 1, Y < 2) = 15 e − 2 x − 3 y dyd x = 5 e − 2 x (e −3 x − e − 6 )d x
0 x 0
1 1
5
= 5 e − 5 x dx − 5e − 6 e − 2 x dx = 1 − e −5 + e − 6 (e − 2 − 1) = 0.9879
0 0
2
2 ∞ 2
b) P(1 < X < 2) = 15 e − 2 x − 3 y dyd x = 5 e −5 x dyd x =(e −5 − e 10 ) = 0.0067
1 x 1

c)
3∞ ∞∞ 3 ∞
− 2 x −3 y − 2 x −3 y −9 −2 x
P (Y > 3) = 15 e dydx + e dydx = 5 e e dx + 5 e −5 x dx
0 3 3 x 0 3

3 5
= − e −15 + e −9 = 0.000308
2 2

5-19
 d)
2 2 2
P ( X < 2, Y < 2) = 15 e − 2 x −3 y dyd x = 5 e − 2 x (e −3 x − e −6 )d x =
0 x 0
2 2
5
( )
= 5 e −5 x dx − 5e −6 e − 2 x dx = 1 − e −10 + e − 6 e − 4 − 1 = 0.9939
2
( )
0 0

∞∞ ∞
1
e) E(X) = 15 xe − 2 x −3 y dyd x = 5 xe −5 x dx = = 0.04
0 x 0 52

f)
∞∞ ∞ ∞
−3 5
E (Y ) = 15 ye − 2 x −3 y dyd x = 5 ye −5 y dy + 3 ye −3 y dy
0 x
2 0 20
3 5 8
=− + =
10 6 15

∞
15 − 2 z − 3 x
5-48 a.) f ( x) = 15 e − 2 x − 3 y dy = (e ) = 5e − 5 x for x > 0
x 3
−5
b) f X (1) = 5e f XY (1, y ) = 15e −2 − 3 y
15e −2 − 3 y
f Y | X =1 ( y ) = = 3e 3− 3 y for 1 <y
5e − 5
∞
∞ ∞
c) E (Y | X = 1) = 3 ye 3−3 y dy = − y e 3−3 y + e 3−3 y dy = 4 / 3
1 1
1
2 15 − 6
d) 3e 3−3 y dy = 1 − e −3 = 0.9502 for 0 < y , f Y (2) = e
1 2
15e −2 x − 6
f X |Y = 2 ( y ) = = 2e − 2 x
15 − 6
e
2

5-20
5-49. The graph of the range of (X, Y) is
y

5
4
3
2
1
0 1 2 3 4 x
1 x +1 4 x +1
cdydx + cdydx = 1
0 0 1 x −1
1 4
= c ( x + 1)dx + 2c dx
0 1
3
= c + 6c = 7.5c = 1
2
Therefore, c = 1/7.5=2/15

0.50.5
1 1
5-50 a) P ( X < 0.5, Y < 0.5) = 7.5
dydx = 30
0 0
0.5x +1 0.5
b) P ( X < 0.5) = 1
7.5
dydx = 1
7.5
( x + 1)dx = 2
15
( 58 ) = 1
12
0 0 0
c)
1 x +1 4 x +1
x x
E( X ) = 7.5
dydx + 7.5
dydx
0 0 1 x −1
1 4
= 1
7.5
( x 2 + x)dx + 2
7.5
( x)dx = 12
15
( 56 ) + 2
7.5
(7.5) = 19
9
0 1
d)
1 x +1 4 x +1
1 1
E (Y ) = 7.5
ydydx + 7.5
ydydx
0 0 1 x −1
1 4
1 ( x +1) 2 ( x +1) 2 − ( x −1) 2
= 7.5 2
dx + 71.5 2
dx
0 1
1 4
1 2 1
= 15
( x + 2 x + 1)dx + 15
4 xdx
0 1

= 151 ( 73 ) + 151 (30 ) = 97

45

5-21
5-51. a. )
x +1
1 x +1
f ( x) = dy = for 0 < x < 1,
0
7 .5 7 .5
x +1
1 x + 1 − ( x − 1) 2
f ( x) = dy = = for 1 < x < 4
x −1
7 .5 7 .5 7 .5
b. )
f XY (1, y ) 1 / 7.5
f Y | X =1 ( y ) = = = 0 .5
f X (1) 2 / 7 .5
f Y | X =1 ( y ) = 0.5 for 0 < y < 2
2 2
y y2
c. ) E (Y | X = 1) = dy = =1
0
2 4 0
0.5 0.5

d.) P (Y < 0.5 | X = 1) = 0.5dy = 0.5 y = 0.25

0 0
5-52 Let X, Y, and Z denote the time until a problem on line 1, 2, and 3, respectively.
a)
P ( X > 40, Y > 40, Z > 40) = [ P ( X > 40)] 3
because the random variables are independent with the same distribution. Now,
∞ ∞
P( X > 40) = 1
40
e− x / 40 dx = −e − x / 40 = e−1 and the answer is
40 40

(e )
−1 3
= e − 3 = 0.0498 .
b) P ( 20 < X < 40,20 < Y < 40,20 < Z < 40) = [ P (20 < X < 40)]3 and
40
P (20 < X < 40) = −e − x / 40 = e − 0.5 − e −1 = 0.2387 .
20
3
The answer is 0.2387 = 0.0136.

c.) The joint density is not needed because the process is represented by three
independent exponential distributions. Therefore, the probabilities may be multiplied.

5-22
5-53 a.) µ=3.2 λ=1/3.2
∞ ∞ x y ∞ x 5
− − − −
P ( X > 5, Y > 5 ) = (1 / 10 .24 ) e 3 .2 3 . 2
dydx = 3 .2 e 3 .2
e 3 .2
dx
5 5 5
5 5
− −
= e 3 .2
e 3 .2
= 0 .0439

∞ ∞ x y ∞ x 10
− − − −
P( X > 10, Y > 10) = (1 / 10.24) e 3.2 3.2
dydx = 3.2 e 3.2
e 3.2
dx
10 10 10
10 10
− −
= e 3.2
e 3.2
= 0.0019

b.) Let X denote the number of orders in a 5-minute interval. Then X is a Poisson
random variable with λ=5/3.2 = 1.5625.

e −1.5625 (1.5625) 2
P ( X = 2) = = 0.256
2!

For both systems, P ( X = 2) P (Y = 2) = 0.256 2 = 0.0655

c. ) The joint probability distribution is not necessary because the two processes are
independent and we can just multiply the probabilities.

5-54 a) fY|X=x(y), for x = 2, 4, 6, 8

3
f(y2)

0 1 2 3 4
y

5-23
 2
b) P(Y < 2 | X = 2) = 2e − 2 y dy = 0.9817
0
∞
c) E (Y | X = 2) = 2 ye − 2 y dy = 1 / 2 (using integration by parts)
0
∞
d) E (Y | X = x) = xye − xy dy = 1 / x (using integration by parts)
0
1 1 − xy f ( x, y )
e) Use fX(x) = = , fY | X ( x, y ) = xe , and the relationship fY | X ( x, y ) = XY
b − a 10 f X ( x)
f XY ( x, y ) xe − xy
Therefore, xe − xy = and f XY ( x, y ) =
1 / 10 10
− xy −10 y −10 y
10 xe 1 − 10 ye −e
f) fY(y) = dx = 2
(using integration by parts)
0 10 10 y

Section 5-4

0.5 1 1 0.5 1 0.5 0.5

5-55. a) P ( X < 0.5) = (8 xyz )dzdydx = (4 xy )dydx = (2 x)dx = x 2 = 0.25
0 0 0 0 0 0 0

b)
0.5 0.5 1
P ( X < 0.5, Y < 0.5) = (8 xyz )dzdydx
0 0 0
0.5 0.5 0.5 0.5
x2
= (4 xy )dydx = (0.5 x)dx = 4
= 0.0625
0 0 0 0

c) P(Z < 2) = 1, because the range of Z is from 0 to 1.

d) P(X < 0.5 or Z < 2) = P(X < 0.5) + P(Z < 2) - P(X < 0.5, Z < 2). Now, P(Z < 2) =1 and
P(X < 0.5, Z < 2) = P(X < 0.5). Therefore, the answer is 1.
1 1 1 1
1
2 x3
e) E ( X ) = (8 x yz )dzdydx = (2 x 2 )dx =
2
3 0 = 2/3
0 0 0 0

5-56 a) P( X < 0.5 Y = 0.5) is the integral of the conditional density f X Y ( x) . Now,
1
f XY ( x,0.5)
f X 0.5 ( x) = and f XY ( x,0.5) = (8 xyz )dz = 4 xy for 0 < x < 1 and
fY (0.5) 0
1 1
0 < y < 1. Also, fY ( y ) = (8 xyz )dzdx = 2 y for 0 < y < 1.
0 0

2x
Therefore, f X 0.5
( x) = = 2 x for 0 < x < 1.
1
0.5
Then, P ( X < 0.5 Y = 0.5) = 2 xdx = 0.25 .
0

5-24
 b) P( X < 0.5, Y < 0.5 Z = 0.8) is the integral of the conditional density of X and Y.
Now, f Z ( z ) = 2 z for 0 < z < 1 as in part a. and
f XYZ ( x, y, z ) 8 xy(0.8)
f XY Z ( x, y ) = = = 4 xy for 0 < x < 1 and 0 < y < 1.
fZ ( z) 2(0.8)
0.50.5 0.5
1
Then, P ( X < 0.5, Y < 0.5 Z = 0.8) = (4 xy )dydx = ( x / 2)dx = 16
= 0.0625
0 0 0
1
5-57. a) fYZ ( y, z ) = (8 xyz )dx = 4 yz for 0 < y < 1 and 0 < z < 1.
0

f XYZ ( x, y, z ) 8 x(0.5)(0.8)
Then, f X YZ ( x) = = = 2 x for 0 < x < 1.
fYZ ( y, z ) 4(0.5)(0.8)
0.5
b) Therefore, P ( X < 0.5 Y = 0.5, Z = 0.8) = 2 xdx = 0.25
0
4
5-58 a) cdzdydx = the volume of a cylinder with a base of radius 2 and a height of 4 =
0
x2 + y2 ≤4

1
(π 2 2 )4 = 16π . Therefore, c =
16π
b) P ( X 2 + Y 2 ≤ 1) equals the volume of a cylinder of radius 2 and a height of 4 (
8π
=8π) times c. Therefore, the answer is = 1 / 2.
16π
c) P(Z < 2) equals half the volume of the region where f XYZ ( x, y, z ) is positive times
1/c. Therefore, the answer is 0.5.
2 4− x2 4 2 4− x 2 2
d) E ( X ) = x
c
dzdydx = 1
c
4 xy dx = 1
c
(8 x 4 − x 2 )dx . Using
2
− 2− 4 − x 2 0 −2 − 4− x −2

3 2
substitution, u = 4 − x 2 , du = -2x dx, and E ( X ) = 1
c
4 u du = −4 2
c 3
(4 − x 2 ) 2 = 0.
−2
4
f ( x,1)
5-59. a) f X 1 ( x) = XY and f XY ( x, y ) = 1
c
dz = 4c = 1
4π
for x 2 + y 2 < 4 .
fY (1) 0
2
4− y 4
Also, f Y ( y ) = 1
c
dzdx = 8c 4 − y 2 for -2 < y < 2.
− 4− y 2 0

Then, f ( x) = 4/c evaluated at y = 1. That is, f X 1 ( x) = 1

for
X y
2
2 3
8
c
4− y
− 3<x< 3.
1
1 1+ 3
Therefore, P ( X < 1 | Y < 1) = 2 3
dx = = 0.7887
− 3
2 3

5-25
 f ( x, y,1) 2 4 − x2 2
b) f XY 1( x, y) = XYZ and f Z ( z) = 1 dydx = 2 4 − x 2 dx
f Z (1) c c
−2 −2
− 4 − x2
Because f Z (z ) is a density over the range 0 < z < 4 that does not depend on Z,
f Z ( z ) =1/4 for
1/ c 1
0 < z < 4. Then, f XY 1 ( x, y ) = = for x 2 + y 2 < 4 .
1 / 4 4π
area in x 2 + y 2 < 1
Then, P ( x 2 + y 2 < 1 | Z = 1) = = 1/ 4 .
4π

f XYZ ( x, y, z )
5-60 f Z xy ( z ) = and from part 5-59 a., f XY ( x, y ) = 1
4π
for x 2 + y 2 < 4 .
f XY ( x, y )
1
16π
Therefore, f Z xy ( z ) = 1
= 1 / 4 for 0 < z < 4.
4π
5-61 Determine c such that f ( xyz ) = c is a joint density probability over the region x>0, y>0
and z>0 with x+y+z<1
1 1− x 1− x − y 1 1− x 1 1− x
y2
f ( xyz ) = c dzdydx = c(1 − x − y )dydx = c( y − xy − ) dx
0 0 0 0 0 0
2 0
1
1
(1 − x) 2 1
(1 − x ) 1 x2 x3
2

= c (1 − x) − x(1 − x) − dx = c dx = c x − +
0
2 0
2 2 2 6 0

1
=c . Therefore, c = 6.
6
1 1− x 1− x − y
5-62 a.) P ( X < 0.5, Y < 0.5, Z < 0.5) = 6 dzdydx The conditions x>0.5, y>0.5,
0 0 0
z>0.5 and x+y+z<1 make a space that is a cube with a volume of 0.125. Therefore the
probability of P ( X < 0.5, Y < 0.5, Z < 0.5) = 6(0.125) = 0.75

b.)
0 . 50 . 5 0 .5
P ( X < 0 .5, Y < 0 .5) = 6 (1 − x − y ) dydx = (6 y − 6 xy − 3 y ) 2 0.5

0
dx
0 0 0
0.5 0 .5
9 9 3
= − 3 x dx = x − x2 = 3/ 4
0
4 4 2 0

c.)
0.51− x 1− x − y 0.51 1− x 0.5 1− x
y2
P( X < 0.5) = 6 dzdydx = 6(1 − x − y )dydx = 6( y − xy − )
0 0 0 0 0 0 2 0
0.5 2
x 1
= 6(
2
− x + )dx = x 3 − 3 x 2 + 3 x
2
( ) 0.5

0
= 0.875
0

5-26
 d. )
1 1− x 1− x − y 1 1− x 1 1− x
y2
E( X ) = 6 xdzdydx = 6 x(1 − x − y )dydx = 6x( y − xy − )
0 0 0 0 0 0 2 0
1
1
x3 x 3x 4 3x 2
= 6( − x 2 + )dx = − 2x 3 + = 0.25
0
2 2 4 2 0

5-63 a.)
1− x
1− x 1− x − y 1− x
y2
f ( x) = 6 dzdy = 6(1 − x − y )dy = 6 y − xy −
0 0 0
2 0
2
x 1
= 6( − x + ) = 3( x − 1) 2 for 0 < x < 1
2 2

b.)
1− x − y

f ( x, y ) = 6 dz = 6(1 − x − y )
0

for x > 0 , y > 0 and x + y < 1

c.)
f ( x, y = 0.5, z = 0,5) 6
f ( x | y = 0.5, z = 0.5) = = = 1 For, x = 0
f ( y = 0.5, z = 0.5) 6

d. ) The marginal f Y ( y ) is similar to f X ( x) and f Y ( y ) = 3(1 − y ) 2 for 0 < y < 1.

f ( x,0.5) 6(0.5 − x)
f X |Y ( x | 0.5) = = = 4(1 − 2 x) for x < 0.5
f Y (0.5) 3(0.25)

5-64 Let X denote the production yield on a day. Then,

1400 −1500
P ( X > 1400) = P ( Z > 10000
) = P( Z > −1) = 0.84134 .
a) Let Y denote the number of days out of five such that the yield exceeds 1400. Then,
by independence, Y has a binomial distribution with n = 5 and p = 0.8413. Therefore,
()
the answer is P (Y = 5) = 55 0.84135 (1 − 0.8413) 0 = 0.4215 .
b) As in part a., the answer is
P (Y ≥ 4) = P (Y = 4) + P (Y = 5)
= ( )0.8413 (1 − 0.8413)
5
4
4 1
+ 0.4215 = 0.8190

5-27
 5-65. a) Let X denote the weight of a brick. Then,
2.75 − 3
P ( X > 2.75) = P ( Z > 0.25
) = P( Z > −1) = 0.84134 .
Let Y denote the number of bricks in the sample of 20 that exceed 2.75 pounds. Then, by
independence, Y has a binomial distribution with n = 20 and p = 0.84134. Therefore,
the answer is P (Y = 20) = 20 ( )
20 0.84134
20
= 0.032 .
b) Let A denote the event that the heaviest brick in the sample exceeds 3.75 pounds.
Then, P(A) = 1 - P(A') and A' is the event that all bricks weigh less than 3.75 pounds. As
in part a., P(X < 3.75) = P(Z < 3) and
P ( A) = 1 − [ P ( Z < 3)] 20 = 1 − 0.99865 20 = 0.0267 .

5-66 a) Let X denote the grams of luminescent ink. Then,

P ( X < 1.14) = P ( Z < 1.140.−31.2 ) = P ( Z < −2) = 0.022750 .
Let Y denote the number of bulbs in the sample of 25 that have less than 1.14 grams.
Then, by independence, Y has a binomial distribution with n = 25 and p =
0.022750. Therefore, the answer is
( )
P (Y ≥ 1) = 1 − P (Y = 0) = 025 0.02275 0 (0.97725) 25 = 1 − 0.5625 = 0.4375 .
b)
P (Y ≤ 5) = P (Y = 0) + P (Y = 1) + P (Y = 2) + ( P (Y = 3) + P (Y = 4) + P(Y = 5)
= ( )0.02275 (0.97725)
25
0
0 25
+ ( )0.02275 (0.97725)
25
1
1 24
+ ( )0.02275
25
2
2
(0.97725) 23
.
+ ( )0.02275 (0.97725)
25
3
3 22
+ ( )0.02275 (0.97725)
25
4
4 21
+ ( )0.02275
25
5
5
(0.97725) 20
= 0.5625 + 0.3274 + 0.09146 + 0.01632 + 0.002090 + 0.0002043 = 0.99997 ≅ 1

c.) P (Y = 0) = ( )0.02275
25
0
0
(0.97725) 25 = 0.5625

d.) The lamps are normally and independently distributed, therefore, the probabilities can
be multiplied.

Section 5-5

5-67. E(X) = 1(3/8)+2(1/2)+4(1/8)=15/8 = 1.875

E(Y) = 3(1/8)+4(1/4)+5(1/2)+6(1/8)=37/8 = 4.625

E(XY) = [1 × 3 × (1/8)] + [1 × 4 × (1/4)] + [2 × 5 × (1/2)] + [4 × 6 × (1/8)]

= 75/8 = 9.375
σ XY = E ( XY ) − E ( X ) E (Y ) = 9.375 − (1.875)(4.625) = 0.703125
V(X) = 12(3/8)+ 22(1/2) +42(1/8)-(15/8)2 = 0.8594
V(Y) = 32(1/8)+ 42(1/4)+ 52(1/2) +62(1/8)-(37/8)2 = 0.7344
σ XY 0.703125
ρ XY = = = 0.8851
σ XσY (0.8594)(0.7344)

5-28
5-68 E ( X ) = −1(1 / 8) + (−0.5)(1 / 4) + 0.5(1 / 2) + 1(1 / 8) = 0.125
E (Y ) = −2(1 / 8) + (−1)(1 / 4) + 1(1 / 2) + 2(1 / 8) = 0.25
E(XY) = [-1× −2 × (1/8)] + [-0.5 × -1× (1/4)] + [0.5 × 1× (1/2)] + [1× 2 × (1/8)] = 0.875
V(X) = 0.4219
V(Y) = 1.6875
σ XY = 0.875 − (0.125)(0.25) = 0.8438
0.8438
ρ XY = σσ σ = X
XY
Y
=1
0.4219 1.6875

5-69.
3 3
c( x + y ) = 36c, c = 1 / 36
x =1 y =1
2
13 13 14 14 13 −1
E( X ) = E (Y ) = E ( XY ) = σ xy = − =
6 6 3 3 6 36
16 16 23
E( X 2 ) = E (Y 2 ) = V ( X ) = V (Y ) =
3 3 36
−1
ρ= 36 = −0.0435
23 23
36 36

5-70 E ( X ) = 0(0.01) + 1(0.99) = 0.99

E (Y ) = 0(0.02) + 1(0.98) = 0.98
E(XY) = [0 × 0 × (0.002)] + [0 × 1× (0.0098)] + [1× 0 × (0.0198)] + [1× 1× (0.9702)] = 0.9702
V(X) = 0.99-0.992=0.0099
V(Y) = 0.98-0.982=0.0196
σ XY = 0.9702 − (0.99)(0.98) = 0
0
ρ XY = σσ σ = XY
X Y
=0
0.0099 0.0196

5-71 E(X1) = np1 = 20(1/3)=6.67

E(X2) = np2=20(1/3)=6.67
V(X1) = np1(1-p1)=20(1/3)(2/3)=4.44
V(X2) = np2(1-p2)=20(1/3)(2/3)=4.44
E(X1X2) =n(n-1)p1p2 =20(19)(1/3)(1/3)=42.22
− 2.267
σ XY = 42.22 − 6.67 2 = −2.267 and ρ XY = = −0.51
(4.44)(4.44)

The sign is negative.

5-29
5-72 From Exercise 5-40, c=8/81.
From Exercise 5-41, E(X) = 12/5, and E(Y) = 8/5
3 x 3 x 3 3
8 8 8 x3 2 8 x5
E ( XY ) = xy ( xy )dyd x = x 2 y 2 dyd x = x dx = dx
81 0 0 81 0 0 81 0 3 81 0 3
8 36
= =4
81 18
12 8
σ xy = 4 − = 0.16
5 5
E( X 2 ) = 6 E (Y 2 ) = 3
V ( x) = 0.24, V (Y ) = 0.44
0.16
ρ= = 0.4924
0.24 0.44

5-73. Similarly to 5-49, c = 2 / 19

1 x +1 5 x +1
2 2
E( X ) = xdydx + xdydx = 2.614
19 0 0
19 1 x −1
1 x +1 5 x +1
2 2
E (Y ) = ydydx + ydydx = 2.649
19 0 0 19 1 x −1
1 x +1 5 x +1
2 2
Now, E ( XY ) = xydydx + xydydx = 8.7763
19 0 0 19 1 x −1
σ xy = 8.7763 − (2.614)(2.649) = 1.85181
E ( X 2 ) = 8.7632 E (Y 2 ) = 9.11403
V ( x) = 1.930, V (Y ) = 2.0968
1.852
ρ= = 0.9206
1.930 2.062

5-30
5-74
1 x +1 4 x +1
x x
E( X ) = 7.5
dydx + 7.5
dydx
0 0 1 x −1
1 4
2
= 1
7.5
( x + x)dx + 2
7.5
( x)dx = 12 (5) +
15 6
2
7.5
(7.5) = 199
0 1
2
E(X )=222,222.2
V(X)=222222.2-(333.33)2=111,113.31
E(Y2)=1,055,556
V(Y)=361,117.11
∞∞
E ( XY ) = 6 × 10 −6 xye −.001x −.002 y dydx = 388,888.9
0 x

σ xy = 388,888.9 − (333.33)(833.33) = 111,115.01

111,115.01
ρ= = 0.5547
111113.31 361117.11

5-75. a) E(X) = 1 E(Y) = 1

∞∞
E ( XY ) = xye − x − y dxdy
0 0
∞ ∞
= xe dx ye − y dy
−x

0 0

= E ( X ) E (Y )
Therefore, σ XY = ρ XY = 0 .

5-76. Suppose the correlation between X and Y is ρ. for constants a, b, c, and d, what is the
correlation between the random variables U = aX+b and V = cY+d?
Now, E(U) = a E(X) + b and E(V) = c E(Y) + d.
Also, U - E(U) = a[ X - E(X) ] and V - E(V) = c[ Y - E(Y) ]. Then,
σ UV = E{[U − E (U )][V − E (V )]} = acE{[ X − E ( X )][Y − E (Y )]} = acσ XY
Also, σ U2 = E[U − E (U )]2 = a 2 E[ X − E ( X )]2 = a 2σ X2 and σ V2 = c 2σ Y2 . Then,

acσ XY ρ XY if a and c are of the same sign

ρ UV = =
a 2σ X2 c 2σ Y2 - ρ XY if a and c differ in sign

5-31
5-77 E ( X ) = −1(1 / 4) + 1(1 / 4) = 0
E (Y ) = −1(1 / 4) + 1(1 / 4) = 0
E(XY) = [-1× 0 × (1/4)] + [-1× 0 × (1/4)] + [1× 0 × (1/4)] + [0 × 1× (1/4)] = 0
V(X) = 1/2
V(Y) = 1/2
σ XY = 0 − (0)(0) = 0
0
ρ XY = σσ σ =
XY
X Y
=0
1/ 2 1/ 2
The correlation is zero, but X and Y are not independent, since, for example, if
y=0, X must be –1 or 1.

5-78 If X and Y are independent, then f XY ( x, y ) = f X ( x) f Y ( y ) and the range of

(X, Y) is rectangular. Therefore,
E ( XY ) = xyf X ( x) fY ( y )dxdy = xf X ( x)dx yfY ( y )dy = E ( X ) E (Y )
hence σXY=0

Section 5-6

5-79 a.)

0.04

0.03

0.02
z(0)
0.01

10
0.00

-2
0 x
-1 0 1 -10
y 2 3 4

5-32
b.)

0.07
0.06
0.05
0.04
0.03
z(.8)
0.02
0.01
10
0.00

-2
0 x
-1 0 1 2 -10
y 3 4

c)

0.07

0.06
0.05

0.04
0.03
z(-.8)
0.02

0.01
10
0.00

-2
0 x
-1 0 1 -10
y 2 3 4

5-33
5-80 Because ρ = 0 and X and Y are normally distributed, X and Y are independent.
Therefore,
P(2.95 < X < 3.05, 7.60 < Y < 7.80) = P(2.95 < X < 3.05) P(7.60 < Y < 7.80) =
P ( 2.095.04− 3 < Z < 3.05 − 3
0.04
) P ( 7.600.−087.70 < Z < 7.80 − 7.70
0.08
) = 0.7887 2 = 0.6220

5-81. Because ρ = 0 and X and Y are normally distributed, X and Y are independent.
Therefore,
µX = 0.1mm σX=0.00031mm µY = 0.23mm σY=0.00017mm
Probability X is within specification limits is
0.099535 − 0.1 0.100465 − 0.1
P (0.099535 < X < 0.100465) = P <Z<
0.00031 0.00031
= P(−1.5 < Z < 1.5) = P ( Z < 1.5) − P( Z < −1.5) = 0.8664
Probability that Y is within specification limits is
0.22966 − 0.23 0.23034 − 0.23
P (0.22966 < X < 0.23034) = P <Z <
0.00017 0.00017
= P (−2 < Z < 2) = P ( Z < 2) − P ( Z < −2) = 0.9545
Probability that a randomly selected lamp is within specification limits is
(0.8664)(.9594)=0.8270

5-82 a) By completing the square in the numerator of the exponent of the bivariate normal PDF, the joint
PDF can be written as
2 2
1 σY x−µ x
( y −( µY + ρ ( x − µ x )) + (1− ρ 2 )
σ Y2 σX σx
−
1 2 (1− ρ 2 )
e
2
f XY ( x , y ) 2πσ x σ y 1 − ρ
f Y |X = x = = 2
f X ( x) x−µ x
σx
1 −
2
e
2π σ x
2
x−µx
σx
Also, fx(x) = 1 −
2 By definition,
e
2π σ x
2 2
1 σY x −µ x
( y − (µY + ρ ( x − µ x )) + (1−ρ 2 )
σY2 σX σx
−
1 2(1− ρ 2 )
e
2
f XY ( x, y ) 2πσ x σ y 1 − ρ
fY | X = x = = 2
f X ( x) x −µ x
σx
−
1 2(1− ρ 2 )
e
2π σ x
2
1 σY
( y − ( µY + ρ ( x −µ x ))
σY2 σX
−
1 2 (1−ρ 2 )
= e
2π σ y 1 − ρ 2

5-34
 Now fY|X=x is in the form of a normal distribution.

σy
b) E(Y|X=x) = µ y + ρ ( x − µ x ) . This answer can be seen from part 5-82a. Since the PDF is in
σx
the form of a normal distribution, then the mean can be obtained from the exponent.

c) V(Y|X=x) = σ 2y (1 − ρ2 ) . This answer can be seen from part 5-82a. Since the PDF is in the form
of a normal distribution, then the variance can be obtained from the exponent.

5-83
∞ ∞ ∞ ∞ ( x−µ X )2 ( y − µY ) 2
− 12 +
f XY ( x, y )dxdy = 1 σ X
2 σ Y
2
dxdy =
2π σ σ e
X Y
− ∞− ∞ −∞−∞

∞ ( x−µ X )2 ∞ ( y − µY ) 2
− 12 − 12
1 σ 2
1 σ 2

σ e dx σ e dy
X Y
2π X 2π Y
−∞ −∞

and each of the last two integrals is recognized as the integral of a normal probability
density function from −∞ to ∞. That is, each integral equals one. Since
f XY( x, y ) = f ( x) f ( y ) then X and Y are independent.

2 2
X −µ X 2ρ ( X −µ X )(Y − µ X ) Y − µY
− +
σX σ X σY σY
−
1 2(1− ρ 2 )
5-84 Let f XY ( x, y ) = e
2πσ xσ y 1 − ρ2
Completing the square in the numerator of the exponent we get:

2 2
2 2
X −µ X 2 ρ ( X − µ X )(Y − µ X ) Y − µY Y − µY X −µ X 2 X −µ X
− + = −ρ + (1 − ρ )
σX σ XσY σY σY σX σX

But,
Y − µY X −µ X 1 σY 1 σY
− ρ = (Y − µ Y ) − ρ ( X −µ x ) = (Y − ( µY + ρ ( X − µ x ))
σY σX σY σX σY σX

Substituting into fXY(x,y), we get

2 2
1 σY x−µx
y − ( µY + ρ ( x − µ x )) + (1− ρ 2 )
σ Y2 σX σx
−
∞ ∞ 1 2(1− ρ 2 )
f XY ( x, y ) = e dydx
−∞ −∞
2πσ xσ y 1 − ρ 2
2
σy
y −( µ y + ρ ( x − µ x ))
σx
2 −
1 x−µx 2σ x2 (1− ρ 2 )
−
∞ 1 2 σx ∞ 1
= e dx × e dy
−∞
2πσ x −∞
2πσ y 1 − ρ 2

The integrand in the second integral above is in the form of a normally distributed random
variable. By definition of the integral over this function, the second integral is equal to 1:

5-35
 2
σy
y −( µ y + ρ ( x − µ x ))
σx
2 −
1 x−µx 2σ x2 (1− ρ 2 )
−
∞ 1 2 σx ∞ 1
e dx × e dy
−∞
2πσ x −∞
2πσ y 1 − ρ 2
2
1 x−µx
−
∞ 1 2 σx
= e dx × 1
−∞
2πσ x

The remaining integral is also the integral of a normally distributed random variable and therefore,
it also integrates to 1, by definition. Therefore,

∞ ∞
f ( x, y ) =1
− ∞ − ∞ XY

5-85
− 0 .5 ( x−µ X )2 2 ρ ( x − µ X )( y − µ Y ) ( y − µY ) 2
∞ − +
1− ρ 2 σ 2 σσ X Y σ 2
f X ( x) = 1
e X Y
dy
πσ σ
2 X Y 1− ρ 2
−∞

2 2
)2 ( y − µY ) ρ ( x−µ X ) ρ ( x−µ X )
− 0.5 ( x − µ X ∞ − 0 .5 − −
1− ρ 2 σ 2 1− ρ 2 σ Y σ X σ X
= 1
σ e
X 1
e dy
2π X 2π σ Y 1− ρ 2
−∞
2

− 0 .5 ( y − µY ) ρ ( x−µ X )
( x−µ X )2 −
− 0 .5
σ 2
∞ 1− ρ 2
σ σ X
= 1
2πσ X
e X 1
e Y
dy
2π σ Y 1− ρ 2
−∞

The last integral is recognized as the integral of a normal probability density with mean
µ Y + σ ρσ( x − µ
Y
X
X ) 2
and variance σ Y (1 − ρ 2 ) . Therefore, the last integral equals one and
the requested result is obtained.

5-36
5-86 E ( X ) = µ X , E (Y ) = µY , V ( X ) = σ X2 , and V (Y ) = σ Y2 . Also,

− 0.5 ( x−µ X )2 2 ρ ( x−µ X )( y − µ Y ) ( y − µ Y ) 2

∞ ∞
1− ρ 2 σ 2
X
−
σσ X Y
+
σ Y
2

( x − µ X )( y − µ Y )e
E ( X − µ X )(Y − µ Y ) = dxdy
− ∞− ∞ 2πσ X σ Y (1 − ρ 2 ) 1 / 2

x−µ X y − µY
Let u = σX
and v = σY
. Then,
−0.5
u 2 − 2 ρuv + v 2
1− ρ 2
∞ ∞
uve
E ( X − µ X )(Y − µ Y ) = 2 1/ 2
σ X σ Y dudv
− ∞− ∞ 2π (1 − ρ )
− 0. 5
[u − ρv ]2 + (1− ρ 2 ) v 2
1− ρ 2
∞ ∞
uve
= σ X σ Y dudv
− ∞− ∞ 2π (1 − ρ 2 ) 1 / 2
The integral with respect to u is recognized as a constant times the mean of a normal
random variable with mean ρv and variance 1 − ρ 2 . Therefore,

∞ ∞
v 2 v2 2
E ( X − µ X )(Y − µY ) = e −0.5v ρvσ X σ Y dv = ρσ X σ Y e −0.5v dv .
−∞ 2π −∞ 2π
The last integral is recognized as the variance of a normal random variable with mean 0
and variance 1. Therefore, E( X − µ X )( Y − µ Y ) = ρσ X σ Y and the correlation between X and
Y is ρ .

Section 5-7

5-87. a) E(2X + 3Y) = 2(0) + 3(10) = 30

b) V(2X + 3Y) = 4V(X) + 9V(Y) = 97
c) 2X + 3Y is normally distributed with mean 30 and variance 97. Therefore,
30 − 30
P(2 X + 3Y < 30) = P( Z < 97
) = P( Z < 0) = 0.5
40 −30
d) P(2 X + 3Y < 40) = P( Z < 97
) = P( Z < 1.02) = 0.8461

5-88 Y = 10X and E(Y) =10E(X) = 50mm.

V(Y)=102V(X)=25mm2

5-37
5-89 a) Let T denote the total thickness. Then, T = X + Y and E(T) = 4 mm,
V(T) = 0.12 + 0.12 = 0.02mm 2 , and σ T = 0.1414 mm.
b)
4 .3 − 4
P (T > 4.3) = P Z > = P ( Z > 2.12)
0.1414
= 1 − P ( Z < 2.12) = 1 − 0.983 = 0.0170

5-90 a) X∼N(0.1, 0.00031) and Y∼N(0.23, 0.00017) Let T denote the total thickness.
Then, T = X + Y and E(T) = 0.33 mm,
V(T) = 0.000312 + 0.00017 2 = 1.25 x10 −7 mm 2 , and σ T = 0.000354 mm.
0.2337 − 0.33
P (T < 0.2337 ) = P Z < = P (Z < −272) ≅ 0
0.000354

b)
0 . 2405 − 0 . 33
P ( T > 0 . 2405 ) = P Z > = P ( Z > − 253 ) = 1 − P ( Z < 253 ) ≅ 1
0 . 000345

5-91. Let D denote the width of the casing minus the width of the door. Then, D is normally
distributed.
a) E(D) = 1/8 V(D) = ( 18 ) 2 + ( 161 ) 2 = 256
5

1 1
−
b) P ( D > 14 ) = P ( Z > 4 8
5
) = P ( Z > 0.89) = 0.187
256

0 − 18
c) P ( D < 0) = P ( Z < 5
) = P ( Z < −0.89) = 0.187
256

5-92 D=A-B-C
a) E(D) = 10 - 2 - 2 = 6 mm
V ( D ) = 0.12 + 0.052 + 0.052 = 0.015mm 2
σ D = 0.1225mm
5 .9 − 6
b) P(D < 5.9) = P(Z < ) = P(Z < -0.82) = 0.206.
0.1225

5-93. a) Let X denote the average fill-volume of 100 cans. σ X = 0.5 2 = 0.05 .
100

12 − 12.1
b) E( X ) = 12.1 and P ( X < 12) = P Z < = P ( Z < −2) = 0.023
0.05
12 − µ
c) P( X < 12) = 0.005 implies that P Z < = 0.005.
0.05
12 − µ
Then 0.05 = -2.58 and µ = 12.129 .
12 − 12.1
d.) P( X < 12) = 0.005 implies that P Z < = 0.005.
σ / 100
Then 12 −12.1
σ / 100
= -2.58 and σ = 0.388 .

5-38
 12 − 12.1
e.) P( X < 12) = 0.01 implies that P Z < = 0.01.
0 .5 / n
Then 12−12.1
0.5 / n
= -2.33 and n = 135.72 ≅ 136 .

5-94 Let X denote the average thickness of 10 wafers. Then, E( X ) = 10 and V( X ) = 0.1.
a) P (9 < X < 11) = P ( 9 −010
.1
< Z < 11−010
.1
) = P (−3.16 < Z < 3.16) = 0.998 .
The answer is 1 − 0.998 = 0.002
b) P( X > 11) = 0.01 and σ X = 1
n
.
Therefore, P ( X > 11) = P ( Z > 11−10
) = 0.01 , 11− 10 = 2.33 and n = 5.43 which is
1 1
n n
rounded up to 6.

c.) P( X > 11) = 0.0005 and σ X = σ 10

.
11−10 11−10
Therefore, P ( X > 11) = P ( Z > σ
) = 0.0005 , σ
= 3.29
10 10

σ = 10 / 3.29 = 0.9612

5-95. X~N(160, 900)

a) Let Y = 25X, E(Y)=25E(X)=4000, V(Y) = 252(900)=562500
P(Y>4300) =
4300 − 4000
P Z> = P ( Z > 0.4) = 1 − P ( Z < 0.4) = 1 − 0.6554 = 0.3446
562500
x − 4000
b.) c) P( Y > x) = 0.0001 implies that P Z > = 0.0001.
562500
x − 4000
Then 750 = 3.72 and x = 6790 .

Supplemental Exercises

5-96 The sum of ( ) ( ) ( ) ( ) ( )

f ( x, y ) = 1 , 1 + 1 + 1 + 1 + 1 = 1
4 8 8 4 4
x y

and f XY ( x, y ) ≥ 0
5-97. a) P( X < 0.5, Y < 1.5) = f XY (0,1) + f XY (0,0) = 1 / 8 + 1 / 4 = 3 / 8 .
b) P( X ≤ 1) = f XY (0,0) + f XY (0,1) + f XY (1,0) + f XY (1,1) = 3 / 4
c) P(Y < 1.5) = f XY (0,0) + f XY (0,1) + f XY (1,0) + f XY (1,1) = 3 / 4
d) P( X > 0.5, Y < 1.5) = f XY (1,0) + f XY (1,1) = 3 / 8
e) E(X) = 0(3/8) + 1(3/8) + 2(1/4) = 7/8.
V(X) = 02(3/8) + 12(3/8) + 22(1/4) - 7/82 =39/64
E(Y) = 1(3/8) + 0(3/8) + 2(1/4) = 7/8.
. V(Y) = 12(3/8) + 02(3/8) + 22(1/4) - 7/82 =39/64

5-39
5-98 a) f X ( x ) = f XY ( x, y ) and f X (0) = 3 / 8, f X (1) = 3 / 8, f X (2) = 1/ 4 .
y

f XY (1, y ) 1/8 1/ 4
b) fY 1 ( y ) = and fY 1 (0) = 3/8
= 1 / 3, fY 1 (1) = 3/8
= 2/3.
f X (1)
c) E (Y | X = 1) = yf XY (1, y ) =0(1/ 3) + 1(2 / 3) = 2 / 3
x =1
d) Because the range of (X, Y) is not rectangular, X and Y are not independent.

e.) E(XY) = 1.25, E(X) = E(Y)= 0.875 V(X) = V(Y) = 0.6094

COV(X,Y)=E(XY)-E(X)E(Y)= 1.25-0.8752=0.4844

0.4844
ρ XY = = 0.7949
0.6094 0.6094
20!
5-99 a. ) P ( X = 2, Y = 4, Z = 14) = 0.10 2 0.20 4 0.7014 = 0.0631
2!4!14!
b.) P ( X = 0) = 0.10 0 0.90 20 = 0.1216
c.) E ( X ) = np1 = 20(0.10) = 2
V ( X ) = np1 (1 − p1 ) = 20(0.10)(0.9) = 1.8
f ( x, z )
d.) f X | Z = z ( X | Z = 19) XZ
f Z ( z)
20!
f XZ ( xz ) = 0.1 x 0.2 20− x − z 0.7 z
x! z!(20 − x − z )!
20!
f Z ( z) = 0.3 20− z 0.7 z
z! (20 − z )!
x 20 − x − z
f XZ ( x, z ) (20 − z )! 0.1 x 0.2 20 − x − z (20 − z )! 1 2
f X | Z = z ( X | Z = 19) = 20 − z
=
f Z ( z) x! (20 − x − z )! 0.3 x! (20 − x − z )! 3 3
Therefore, X is a binomial random variable with n=20-z and p=1/3. When z=19,
2 1
f X |19 (0) = and f X |19 (1) = .
3 3
2 1 1
e.) E ( X | Z = 19) = 0 +1 =
3 3 3
5-100 Let X, Y, and Z denote the number of bolts rated high, moderate, and low.
Then, X, Y, and Z have a multinomial distribution.
20!
a) P ( X = 12, Y = 6, Z = 2) = 0.6120.360.12 = 0.0560 .
12!6!2!
b) Because X, Y, and Z are multinomial, the marginal distribution of Z is binomial with
n = 20 and p = 0.1.
c) E(Z) = np = 20(0.1) = 2.

5-40
 f XZ (16, z ) 20!
5-101. a) f Z |16 ( z ) = and f XZ ( x, z ) = 0.6 x 0.3 ( 20 − x − z ) 0.1 z for
f X (16) x! z! (20 − x − z )!
x + z ≤ 20 and 0 ≤ x,0 ≤ z . Then,
20!
0.616 0.3( 4 − z ) 0.1z
f Z 16 ( z ) =
16!z!( 4 − z )!
20! 16 4
= 4!
( ) ( )
0.3 4 − z 0.1 z
z!( 4 − z )! 0.4 0.4
16!4! 0.6 0.4
for 0 ≤ z ≤ 4 . That is the distribution of Z given X = 16 is binomial with n = 4 and
p = 0.25.
b) From part a., E(Z) = 4 (0.25) = 1.
c) Because the conditional distribution of Z given X = 16 does not equal the marginal
distribution of Z, X
and Z are not independent.

5-102 Let X, Y, and Z denote the number of calls answered in two rings or less, three or
four rings, and five rings or more, respectively.
10!
a) P ( X = 8, Y = 1, Z = 1) = 0.780.2510.051 = 0.0649
8!1!1!
b) Let W denote the number of calls answered in four rings or less. Then, W is a
binomial random variable with n = 10 and p = 0.95.
Therefore, P(W = 10) = 10 10 0
( )
10 0.95 0.05 = 0.5987 .
c) E(W) = 10(0.95) = 9.5.

f XZ (8, z ) 10!
5-103 a) f Z 8 ( z ) = and f XZ ( x, z ) = 0.70 x 0.25(10 − x − z ) 0.05 z for
f X (8) x! z!(10 − x − z )!
x + z ≤ 10 and 0 ≤ x,0 ≤ z . Then,
10!
0.70 8 0.25 ( 2 − z ) 0.05 z
f Z 8 ( z) =
8!z!( 2 − z )!
10! 8 2
= 2!
( ) ( )
0.25 2 − z 0.05 z
z !( 2 − z )! 0.30 0.30
8!2!
0.70 0.30
for 0 ≤ z ≤ 2 . That is Z is binomial with n =2 and p = 0.05/0.30 = 1/6.
b) E(Z) given X = 8 is 2(1/6) = 1/3.
c) Because the conditional distribution of Z given X = 8 does not equal the marginal
distribution of Z, X and Z are not independent.

3 2 3 2 3
y2 3
5-104 cx 2 ydydx = cx 2 2
dx = 2c x3 = 18c . Therefore, c = 1/18.
0 0 0 0 0

5-41
 1 1 1 1 1
y2 1 x3
5-105. a) P ( X < 1, Y < 1) = 1
18
x 2 ydydx = 1
18
x2 2
dx = 36 3
= 1
108
0 0 0 0 0

2.5 2 2.5 2 2.5

y2 1 x3
b) P ( X < 2.5) = 1
18
x 2 ydydx = 1
18
x2 2
dx = 9 3
= 0.5787
0 0 0 0 0

3 2 3 2 3
y2 1 x3
c) P (1 < Y < 2.5) = 1
18
x 2 ydydx = 1
18
x2 2
dx = 12 3
= 3
4
0 1 0 1 0

d)
3 1.5 3 1.5 3
2 y2 5 x3
P ( X > 2,1 < Y < 1.5) = 1
18
x ydydx = 1
18
x2 2
dx = 144 3
2 1 2 1 2

95
= 432
= 0.2199
3 2 3 3
3 1 x4
e) E ( X ) = 1
18
x ydydx = 1
18
x 3 2dx = 9 4
= 9
4
0 0 0 0

3 2 3 3
4 x3
f) E (Y ) = 1
18
x 2 y 2 dydx = 1
18
x 2 83 dx = 27 3
= 4
3
0 0 0 0

2
5-106 a) f X ( x) = 1
18
x 2 ydy = 19 x 2 for 0 < x < 3
0
1
f XY (1, y ) 18
y y
b) fY X ( y ) = = 1
= for 0 < y < 2.
f X (1) 9
2
2 3
f XY ( x,1) 181 x y
c) f X 1 ( x) = = and fY ( y ) = 1
18
x 2 ydx = 2
for 0 < y < 2 .
fY (1) fY (1) 0
1
18
x2 1 2
Therefore, f X 1 ( x) = = x for 0 < x < 3.
1/ 2 9

5-107. The region x2 + y 2 ≤ 1 and 0 < z < 4 is a cylinder of radius 1 ( and base area π ) and
1
height 4. Therefore, the volume of the cylinder is 4 π and f XYZ ( x, y, z) = for x2 + y 2 ≤ 1
4π
and 0 < z < 4.
a) The region X 2 + Y 2 ≤ 0.5 is a cylinder of radius 0.5 and height 4. Therefore,
4 ( 0.5π )
P( X 2 + Y 2 ≤ 0.5) = 4π
= 1/ 2 .
b) The region X 2 + Y 2 ≤ 0.5 and 0 < z < 2 is a cylinder of radius 0.5 and height 2.
Therefore,
2 ( 0.5π )
P( X 2 + Y 2 ≤ 0.5, Z < 2) = 4π
= 1/ 4

5-42
 f XYZ ( x, y,1)
c) f XY 1 ( x, y ) = and f Z ( z ) = 1
4π dydx = 1 / 4
f Z (1) x 2 + y 2 ≤1

1 / 4π
for 0 < z < 4. Then, f XY 1 ( x, y ) = = 1
π
for x 2 + y 2 ≤ 1 .
1/ 4
4 1− x 2 4
d) f X ( x) = 1
4π
dydz = 1
2π
1 − x 2 dz = π2 1 − x 2 for -1 < x < 1
0 − 1− x 2 0

4
f XYZ (0,0, z ) 2 2
and f XY ( x, y ) = 4π dz = 1 / π for x + y ≤ 1 . Then,
1
5-108 a) f Z 0, 0 ( z ) =
f XY (0,0) 0

1 / 4π
f Z 0,0 ( z ) = = 1 / 4 for 0 < z < 4 and µ Z 0, 0 = 2 .
1/ π
f ( x, y, z ) 1 / 4π
b) f Z xy ( z ) = XYZ = = 1 / 4 for 0 < z < 4. Then, E(Z) given X = x and Y = y is
f XY ( x, y ) 1/ π
4
z
4
dz = 2 .
0

1 1
5-109. f XY ( x, y ) = c for 0 < x < 1 and 0 < y < 1. Then, cdxdy = 1 and c = 1. Because
0 0

f XY ( x, y ) is constant, P( X − Y < 0.5) is the area of the shaded region below

1
0.5

0 0.5 1

That is, P( X − Y < 0.5) = 3/4.

5-110 a) Let X1, X 2 ,..., X 6 denote the lifetimes of the six components, respectively. Because of
independence,
P( X1 > 5000, X 2 > 5000,..., X 6 > 5000 ) = P( X1 > 5000)P( X 2 > 5000 )... P( X 6 > 5000)

If X is exponentially distributed with mean θ , then λ = 1 and

θ
∞ ∞
P( X > x) = 1
θ
e − t / θ dt = −e − t / θ = e − x / θ . Therefore, the answer is
x x

e −5 / 8 e −0.5 e −0.5 e −0.25 e −0.25 e −0.2 = e −2.325 = 0.0978 .

5-43
 b) The probability that at least one component lifetime exceeds 25,000 hours is the
same as 1 minus the probability that none of the component lifetimes exceed
25,000 hours. Thus,
1-P(Xa<25,000, X2<25,000, …, X6<25,000)=1-P(X1<25,000)…P(X6<25,000)
=1-(1-e-25/8)(1-e-2.5)(1-e-2.5)(1-e-1.25)(1-e-1.25)(1-e-1)=1-.2592=0.7408

5-111. Let X, Y, and Z denote the number of problems that result in functional, minor, and no
defects, respectively.
a) P ( X = 2, Y = 5) = P ( X = 2, Y = 5, Z = 3) = 210 !
!5!3!
0.2 2 0.5 5 0.3 3 = 0.085
b) Z is binomial with n = 10 and p = 0.3.
c) E(Z) = 10(0.3) = 3.

5-112 a) Let X denote the mean weight of the 25 bricks in the sample. Then, E( X ) = 3 and
σ X = 0.25
25
= 0.05 . Then, P( X < 2.95) = P(Z < 2 .095 − 3 ) = P (Z < -1) = 0.159.
. 05
x−3 x−3
b) P( X > x) = P( Z > ) = 0.99 . So, = -2.33 and x=2.8835.
.05 0.05

5-113. a.)
18.25 5.25
Because cdydx = 0.25c, c = 4. The area of a panel is XY and P(XY > 90) is
17.75 4.75

the shaded area times 4 below,

5.25
4.75

17.25 18.25
18.25 5.25 18.25 18.25
That is, 4dydx = 4 5.25 − 90x dx = 4(5.25 x − 90 ln x ) = 0.499
17.75 90 / x 17.75 17.75

b. The perimeter of a panel is 2X+2Y and we want P(2X+2Y >46)

18.25 5.25 18.25
4dydx = 4 5.25 − (23 − x)dx
17.75 23− x 17.75
18.25
x2 18.25
= 4 (−17.75 + x)dx = 4(−17.75 x + ) = 0 .5
17.75
2 17.75

5-44
5-114 a)Let X denote the weight of a piece of candy and X∼N(0.1, 0.01). Each package has 16
candies, then P is the total weight of the package with 16 pieces and E( P ) = 16(0.1)=1.6
ounces and V(P) = 162(0.012)=0.0256 ounces2
b) P ( P < 1.6) = P ( Z < 1.06.−161.6 ) = P ( Z < 0) = 0.5 .
c) Let Y equal the total weight of the package with 17 pieces, E(Y) = 17(0.1)=1.7
ounces and V(Y) = 172(0.012)=0.0289 ounces2
−1.7
P(Y < 1.6) = P( Z < 1.06.0289 ) = P( Z < −0.59) = 0.2776 .
5-115. Let X denote the average time to locate 10 parts. Then, E( X ) =45 and σ X = 30
10
60 − 45
a) P ( X > 60) = P ( Z > 30 / 10
) = P ( Z > 1.58) = 0.057
b) Let Y denote the total time to locate 10 parts. Then, Y > 600 if and only if X > 60.
Therefore, the answer is the same as part a.

5-116 a) Let Y denote the weight of an assembly. Then, E(Y) = 4 + 5.5 + 10 + 8 = 27.5 and
V(Y)= 0.4 2 + 0.5 2 + 0.2 2 + 0.5 2 = 0.7 .
29.5 − 27.5
P (Y > 29.5) = P( Z > 0.7
) = P ( Z > 2.39) = 0.0084
b) Let X denote the mean weight of 8 independent assemblies. Then, E( X ) = 27.5 and
V( X ) = 0.7/8 = 0.0875. Also, P ( X > 29) = P ( Z > 290−.0875
27.5
) = P ( Z > 5.07) = 0 .

5-117

0.07

0.06
0.05

0.04
0.03
z(-.8)
0.02

0.01
10
0.00

-2
0 x
-1 0 1 -10
y 2 3 4

5-45
5-118
−1
1 {( x −1) 2 −1 .6 ( x −1)( y − 2 ) + ( y − 2 ) 2 }
0.72
f XY ( x , y ) = e
1 .2π
−1
{( x −1) 2 −1.6 ( x −1)( y − 2 ) + ( y − 2 ) 2 }
1 2 ( 0 .36 )
f XY ( x , y ) = e
2π .36
−1
{( x −1) 2 − 2 (. 8 )( x −1)( y − 2 ) + ( y − 2 ) 2 }
1 2 (1−.8 2 )
f XY ( x , y ) = e
2π 1 − .8 2
E ( X ) = 1 , E (Y ) = 2 V ( X ) = 1 V (Y ) = 1 and ρ = 0.8

5-119 Let T denote the total thickness. Then, T = X1 + X2 and

a.) E(T) = 0.5+1=1.5 mm
V(T)=V(X1) +V(X2) + 2Cov(X1X2)=0.01+0.04+2(0.014)=0.078mm2
where Cov(XY)=ρσXσY=0.7(0.1)(0.2)=0.014

1 − 1 .5
b.) P (T < 1) = P Z < = P ( Z < −1.79) = 0.0367
0.078
c.) Let P denote the total thickness. Then, P = 2X1 +3 X2 and
E(P) =2(0.5)+3(1)=4 mm
V(P)=4V(X1) +9V(X2) +
2(2)(3)Cov(X1X2)=4(0.01)+9(0.04)+2(2)(3)(0.014)=0.568mm2
where Cov(XY)=ρσXσY=0.7(0.1)(0.2)=0.014

5-120 Let T denote the total thickness. Then, T = X1 + X2 + X3 and

a) E(T) = 0.5+1+1.5 =3 mm
V(T)=V(X1) +V(X2) +V(X3)+2Cov(X1X2)+ 2Cov(X2X3)+
2Cov(X1X3)=0.01+0.04+0.09+2(0.014)+2(0.03)+ 2(0.009)=0.246mm2
where Cov(XY)=ρσXσY

1 .5 − 3
b.) P (T < 1.5) = P Z < = P ( Z < −6.10) ≅ 0
0.246

5-46
5-121 Let X and Y denote the percentage returns for security one and two respectively.
If ½ of the total dollars is invested in each then ½X+ ½Y is the percentage return.
E(½X+ ½Y)= 0.05 (or 5 if given in terms of percent)
V(½X+ ½Y)=1/4 V(X)+1/4V(Y)+2(1/2)(1/2)Cov(X,Y)
where Cov(XY)=ρσXσY=-0.5(2)(4)=-4
V(½X+ ½Y)=1/4(4)+1/4(6)-2=3
Also, E(X)=5 and V(X) = 4. Therefore, the strategy that splits between the securities has a lower
standard deviation of percentage return than investing 2million in the first security.

Mind-Expanding Exercises

5-122. By the independence,

P( X 1 ∈ A1 , X 2 ∈ A2 ,..., X p ∈ A p ) = ... f X 1 ( x1 ) f X 2 ( x 2 )... f X p ( x p )dx1 dx 2 ...dx p

A1 A2 Ap

= f X 1 ( x1 )dx1 f X 2 ( x 2 )dx 2 ... f X p ( x p )dx p

A1 A2 Ap

= P( X 1 ∈ A1 ) P( X 2 ∈ A2 )...P( X p ∈ A p )

5-123 E (Y ) = c1µ1 + c2 µ 2 + ... + c p µ p .

Also,
V (Y ) = [c x + c x + ... + c x − (c µ + c µ + ... + c µ )]
1 1 2 2 p p 1 1 2 2 p p
2
f X 1 ( x1 ) f X 2 ( x2 )... f X p ( x p )dx1dx2 ...dx p
= [c ( x − µ ) + ... + c ( x − µ )] f ( x ) f ( x )... f
1 1 1 p p p
2
X1 1 X2 2 Xp ( x p )dx1dx2 ...dx p

Now, the cross-term

c1c ( x1 − µ1 )( x2 − µ 2 ) f X 1 ( x1 ) f X 2 ( x2 )... f X p ( x p )dx1dx2 ...dx p
2

[
= c1c2 ( x1 − µ1 ) f X 1 ( x1 )dx1 ][ ( x 2 − µ 2 ) f X 2 ( x2 )dx2 = 0 ]
from the definition of the mean. Therefore, each cross-term in the last integral for V(Y)
is zero and
V (Y ) = [ c (x − µ )
2
1 1 1
2
][
f X 1 ( x1 )dx1 ... c 2p ( x p − µ p ) 2 f X p ( x p )dx p ]
= c12V ( X 1 ) + ... + c 2pV ( X p ).

5-47
 a b a b
5-124 f XY ( x, y )dydx = cdydx = cab . Therefore, c = 1/ab. Then,
0 0 0 0
b
a
1
f X ( x) = cdy = a
for 0 < x < a, and f Y ( y) = cdx = 1 for 0 < y < b. Therefore,
b
0 0

f XY (x,y)=f X (x)f Y (y) for all x and y and X and Y are independent.

b b b
5-125 f X ( x) = g ( x)h( y )dy = g ( x) h( y )dy = kg ( x) where k = h( y )dy. Also,
0 0 0
a
f Y ( y ) = lh( y ) where l = g ( x)dx. Because f XY (x,y) is a probability density
0
a b a b
function, g ( x)h( y )dydx = g ( x)dx h( y )dy = 1. Therefore, kl = 1 and
0 0 0 0

f XY (x,y)=f X (x)f Y (y) for all x and y.

Section 5-8 on CD

1
S5-1. f Y ( y) = at y = 3, 5, 7, 9 from Theorem S5-1.
4

S5-2. Because X ≥ 0 , the transformation is one-to-one; that is y = x2 and x= y . From Theorem S5-2,

f Y ( y) = f X ( y ) = ( )p (1 − p)3
y
y 3− y
for y = 0, 1, 4, 9.

If p = 0.25, fY ( y ) = ( )(0.25) (0.75)

3
y
y 3− y
for y = 0, 1, 4, 9.

y − 10 1 y − 10
S5-3. a) fY ( y ) = f X = for 10 ≤ y ≤ 22
2 2 72
22
y 2 − 10 y
b) E (Y ) =
72
dy =
1
72
( y3
3 − 102y
2
) 22

10
= 18
10

y −y −y −y −y
−2
S5-4. Because y = -2 ln x, e = x . Then, f Y ( y ) = f X (e 2 ) − 12 e 2
= 12 e 2 for 0 ≤ e 2 ≤ 1 or
y ≥ 0 , which is an exponential distribution (which equals a chi-square distribution with k = 2 degrees of
freedom).

5-48
S5-5. a) Let Q = R. Then,
p
p = i 2r i= q
and
q=r r=q
∂i ∂i
∂p ∂q
1
( pq ) −1 / 2 − 12 p 1 / 2 q −3 / 2 1
J= ∂r ∂r
= 2
= 2 ( pq ) −1 / 2
∂p ∂q 0 1
f PQ ( p, q ) = f IR ( p
q
, q ) 12 ( pq ) −1 / 2 = 2 ( ) p
q
1
2
( pq ) −1 / 2 = q −1
p
for 0 ≤ q
≤ 1, 0 ≤ q ≤ 1
That is, for 0 ≤ p ≤ q, 0 < q ≤ 1 .
1
f P ( p ) = q −1dq = − ln p for 0 < p ≤ 1.
p
1
b) E ( P ) = − p ln p dp . Let u = ln p and dv = p dp. Then, du = 1/p and
0

2 1
p p2
1
p p2
1 1
v= . Therefore, E ( P) = − (ln p ) 2
+ 2
dp = 4
=
2 0
0
0 4

− 12 e− y
S5-6. a) If y = x 2 , then x = y for x ≥ 0 and y ≥ 0 . Thus, f Y ( y ) = f X ( y ) 12 y = for
2 y
y > 0.
2
b) If y = x 1 / 2 , then x = y 2 for x≥0 and y ≥ 0 . Thus, fY ( y ) = f X ( y 2 )2 y = 2 ye− y for y >
0.
y y
c) If y = ln x, then x = e y for x ≥ 0 . Thus, f Y ( y ) = f X (e y )e y = e y e − e = e y − e for
−∞ < y < ∞.

∞ ∞ ∞
S5-7. a) Now, av 2 e −bv dv must equal one. Let u = bv, then 1 = a ( ub ) 2 e −u du = a3 u 2 e −u du. From
0 0
b b 0
a 2a b3
the definition of the gamma function the last expression is Γ (3) = 3 . Therefore, a = .
b3 b 2
mv 2 2w
b) If w= , then v= for v ≥ 0, w ≥ 0 .
2 m
b 3 2 w −b 2mw
f W ( w) = f V ( ) 2 w dv
m
dw
=
2m
e (2mw) −1 / 2

b 3 m −3 / 2 −b 2w

= w1 / 2 e m

2
for w ≥ 0.

5-49
 1 1
S5-8. If y = e x , then x = ln y for 1 ≤ x ≤ 2 and e 1 ≤ y ≤ e 2 . Thus, fY ( y ) = f X (ln y ) = for
y y
1
1≤ ln y ≤ 2 . That is, fY ( y ) = for e ≤ y ≤ e2 .
y

∞
S5-9. Now P (Y ≤ a ) = P ( X ≥ u (a )) = f X ( x)dx . By changing the variable of integration from x to y
u(a)
−∞
by using x = u(y), we obtain P(Y ≤ a ) = f X (u ( y ))u ' ( y )dy because as x tends to ∞, y = h(x) tends
a
a
to - ∞. Then, P (Y ≤ a ) = f X (u ( y ))(−u ' ( y ))dy . Because h(x) is decreasing, u'( y) is negative.
−∞
Therefore, | u'( y) | = - u'( y) and Theorem S5-1 holds in this case also.

S5-10. If y = ( x − 2) 2 , then x = 2 − y for 0≤ x≤2 and x = 2+ y for 2 ≤ x ≤ 4 . Thus,

fY ( y ) = f X (2 − y ) | − 12 y −1 / 2 | + f X (2 + y ) | 12 y −1 / 2 |
2− y 2+ y
= +
16 y 16 y
= ( 14 ) y −1 / 2 for 0 ≤ y ≤ 4

a
S5-11. a) Let a = s1s 2 and y = s1 . Then, s1 = y, s 2 = and
y
∂s ∂s 1 1
0 1
∂a ∂y 1 a 1
J= = − y −3 / 2 = − y . Then,
∂s ∂s 2
y
2
2
∂a ∂y
f AY (a, y ) = f S S ( y, ay )( 1y ) = 2 y ( 8ay )( 1y ) = 4ay for
1 2
0 ≤ y ≤ 1 and 0 ≤ a ≤ 4 . That is, for
y
0 ≤ y ≤ 1 and 0 ≤ a ≤ 4 y .
1
a a
b) f A (a) = dy = − ln( a4 ) for 0 < a ≤ 4.
a /4 4y 4

5-50
S5-12. Let r = v/i and s = i. Then, i = s and v = rs
∂i ∂i
J = ∂r ∂s = 0 1 = s
∂v ∂v s r
∂r ∂s
f RS (r , s ) = f IV ( s, rs ) s = e −rs s for rs ≥ 0 and 1 ≤ s ≤ 2 . That is,
− rs
f RS (r , s) = se for 1 ≤ s ≤ 2 and r ≥ 0 .
2
− rs − rs − e − rs
Then, f R (r ) = se ds . Let u = s and dv = e ds. Then, du = ds and v =
1
r
Then,
2 2 2
−e−rs −e−rs e−r −2e−2r e−rs
f R (r) = −s + ds= − 2
r 1 1 r r r 1
e−r −2e−2r e−r −e−2r
= +
r r2
e−r (r +1) −e−2r (2r +1)
=
r2
for r > 0.

Section 5-9 on CD

m
e tx 1 m
(e t ) m +1 − e t e t (1 − e tm )
S5-13 . a) E (e tX ) = = (e t ) x = =
x =1 m m x =1 m(e t − 1) m(1 − e t )
1 t
b) M (t ) = e (1 − e tm )(1 − e t ) −1 and
m
dM(t ) 1 t
dt
{
= e (1 − etm )(1 − et )−1 + et (−metm )(1 − et )−1 + et (1 − etm )(−1)(1 − et )−2 (−et )
m
}
dM (t ) et tm tm (1 − e tm )e t
= 1 − e − me +
dt m(1 − e t ) 1− et
et
= {
1 − e tm − me tm + me ( m +1) t }
m(1 − e t ) 2
Using L’Hospital’s rule,
dM (t ) et − me tm − m 2 e tm + m(m + 1)e ( m +1)t
lim = lim lim
t →0 dt t →0 m t →0 − 2(1 − e t )e t
et − m 2 e tm − m 3 e tm + m(m + 1) 2 e ( m +1)t
= lim lim
t →0 m t →0 − 2(1 − e t )e t − 2e t (−e t )
1 m(m + 1) 2 − m 2 − m 3 m 2 + m m + 1
= × = =
m 2 2m 2

5-51
 m +1
Therefore, E(X) = .
2
d 2 M (t ) d 2 m
1 1 m
d2 (tx ) 2
= 2 e tx = 1 + tx + +
dt 2 dt x =1 m m x =1 dt
2
2
m
1
= ( x 2 + term sin volvingpow ersoft )
m x =1
Thus,

d 2 M (t ) 1 m(m + 1)(m + 2) (m + 1)(2m + 2)

= =
dt 2 t =0 m 6 6

Then,
2m 2 + 3m + 1 (m + 1) 2 4m 2 + 6m + 2 − 3m 2 − 6m − 3
V (X ) = − =
6 4 12
2
m −1
=
12

∞
e −λ λx ∞
( λe t ) x t t
S5-14. a) E (e tX ) = e tx = e −λ = e −λ e λe = e λ ( e −1)
x =0 x! x =0 x!
dM (t ) t
b) = λe t e λ ( e −1)
dt
dM (t )
= λ = E( X )
dt t = 0
d 2 M (t ) 2 2 t λ ( e t −1) t λ ( e t −1)
= λ e e + λ e e
dt 2
d 2 M (t )
2
= λ2 + λ
dt t =0

V ( X ) = λ 2 + λ − λ2 = λ

5-52
 ∞
p ∞ t
S5-15 a) E(etX ) = etx (1 − p)x−1 p = [e (1 − p)]x
x =1 1 − p x=1
et (1 − p ) p pet
= =
1 − (1 − p )et 1 − p 1 − (1 − p )et
dM (t )
b) = pe t (1 − (1 − p ) e t ) − 2 (1 − p ) e t + pe t (1 − (1 − p ) e t ) −1
dt
= p (1 − p )e 2t (1 − (1 − p )et )−2 + pet (1 − (1 − p )et )−1
dM(t) 1− p 1
= +1= = E(X )
dt t =0 p p
d 2M(t)
= p(1 − p)e2t 2(1− (1 − p)et )−3(1− p)et + p(1− p)(1 − (1 − p)et )−2 2e2t
dt2
+ pet (1− (1− p)et )−2 (1− p)et + pet (1− (1− p)et )−1
d 2 M (t ) 2(1 − p ) 2 2(1 − p ) 1 − p 2(1 − p ) 2 + 3 p (1 − p ) + p 2
= + + + 1 =
dt 2 t = 0 p2 p p p2
2 + 2 p2 − 4 p + 3 p − 3 p2 + p2 2 − p
= =
p2 p2
2 − p 1 1− p
V (X ) = − 2 = 2
p2 p p

S5-16. M Y (t ) = Ee tY = Ee t ( X 1 + X 2 ) = Ee tX 1 Ee tX 2
= (1 − 2t ) − k1 / 2 (1 − 2t ) − k 2 / 2 = (1 − 2t ) − ( k1 + k 2 ) / 2
Therefore, Y has a chi-square distribution with k 1 + k 2 degrees of freedom.
∞ ∞
S5-17. a) E (e tX ) = e tx 4 xe − 2 x dx = 4 xe ( t − 2) x dx
0 0

Using integration by parts with u = x and dv = e (t − 2 ) x dx and du = dx,

(t − 2) x
e
v= we obtain
t−2
∞ ∞ ∞ ∞
xe (t − 2) x e (t −2 ) x xe ( t − 2) x e (t −2 ) x
4 − dx = 4 −
t−2 0 0
t−2 t−2 0
(t − 2) 2 0

4
This integral only exists for t < 2. In that case, E (e tX ) = for t < 2
(t − 2) 2
dM (t ) dM (t )
b) = −8(t − 2) −3 and = −8(−2) −3 = 1 = E ( X )
dt dt t = 0
d 2 M (t ) d 2 M (t ) 24 3 3 1
c) 2
= 24(t − 2) − 4 and 2
= = . Therefore, V ( X ) = − 12 =
dt dt t =0
16 2 2 2

5-53
 β β
tX etx etx etβ − etα
S5-18. a) E (e ) = dx = =
α β −α t(β − α ) α t(β − α )
dM (t ) e tβ − e tα β e tβ − αe tα
b) = +
dt − ( β − α )t 2 t (β − α )
( βt − 1)e tβ − (αt − 1)e tα
=
t 2 (β − α )
Using L’Hospital’s rule,
dM (t ) (βt − 1)βe tβ + βe tβ − (αt − 1)αe tα − αe tα
lim =
t →0 dt 2t (β − α )
dM (t ) β 2 (βt − 1)e tβ + β 2 e tβ + β 2 e tβ − α 2 (αt − 1)e tα − α 2 e tα − α 2 e tα
lim =
t →0 dt 2(β − α )
dM (t ) β 2 − α 2 (β + α )
lim = = = E( X )
t →0 dt 2(β − α ) 2
b
d 2 M (t ) d 2 1 tx 1 d 2 e tb − e ta
= e dx = ( )
dt 2 dt 2 a b − a b − a dt 2 t
(tb) 2 (tb) 3 (ta ) 2 (ta ) 3
tb + + − ta − − + ...
1 d2 2 3! 2 3!
= =
b − a dt 2 t

1 d2 (b 2 − a 2 )t (b 3 − a 3 )t 2
= b − a + + + ...
b − a dt 2 2 3!
b 3 − a 3 b 2 + ba + a 2
= =
3(b − a ) 3
Thus,
b 2 + ba + a 2 (b + a ) 2 b 2 − 2ab + a 2 (b − a ) 2
V(X)= − = =
3 4 12 12

5-54
 ∞ ∞
S5-19. a) M (t ) = e tx λe − λx dx = λ e ( t − λ ) x dx
0 0

(t −λ ) x ∞
e −λ 1
=λ = = = (1 − t
λ
)−1 for t < λ
t −λ 0
t − λ 1 − λt
dM (t ) 1
= (−1)(1 − λt ) ( −λ1 ) =
−2
b)
λ (1 − λt )
2
dt
dM (t ) 1
=
dt t =0 λ
d 2 M (t ) 2
2
= 2
dt λ (1 − λt ) 3
d 2 M (t ) 2
=
dt 2 t =0
λ2
2
2 1 1
V (X ) = 2
− =
λ λ λ2
∞ ∞
λ λr
S5-20. a) M (t ) = e tx (λx) r −1 e − λx dx = x r −1 e ( t − λ ) x dx
0
Γ (r ) Γ(r ) 0
Let u = (λ-t)x. Then,
∞ r −1
λr u −u du λr Γ(r ) 1
M (t ) = e = r
= t r
= (1 − λt ) − r from
Γ( r ) 0 λ −t λ − t Γ(r )(λ − t ) (1 − λ )
the definition of the gamma function for t < λ.
b) M ' ( t ) = − r (1 − t ) − r − 1 ( − 1 )
λ λ
r
M ' (t ) t = 0 = = E( X )
λ
r (r + 1)
M ' ' (t ) = (1 − λt ) r − 2
λ2
r (r + 1)
M ' ' (t ) t = 0 =
λ2
2
r (r + 1) r r
V (X ) = 2
− =
λ λ λ2

5-55
 n
E (e ) = ∏ E (e tX i ) = (1 − λt )
tY −n
S5-21. a)
i =1
b) From Exercise S5-20, Y has a gamma distribution with parameter λ and n.

2 2 2
µ1t +σ 12 t2 + µ 2t +σ 22 t2 ( µ1 + µ 2 ) t + (σ 12 +σ 22 ) t2
S5-22. a) M Y (t ) = e =e
b) Y has a normal distribution with mean µ 1 + µ 2 and variances σ 12 + σ 22

1 k
S5-23. Because a chi-square distribution is a special case of the gamma distribution with λ = and r = , from
2 2
Exercise S5-20.
M (t ) = (1 − 2t ) − k / 2
k − k −1 − k −1
M ' (t ) = − (1 − 2t ) 2 (−2) = k (1 − 2t ) 2
2
M ' (t ) t =0 = k = E( X )
− k2 − 2
M ' ' (t ) = 2k ( k2 + 1)(1 − 2t )
M ' ' (t ) t = 0 = 2k ( k2 + 1) = k 2 + 2k
V ( X ) = k 2 + 2k − k 2 = 2k
2 r
S5-24. a) M (t ) = M (0) + M ' (0)t + M ' (0) t2! + ... + M ( r ) (0) tr! + ... by Taylor’s expansion. Now, M(0)
= 1 and M ( r ) (0) = µ r' and the result is obtained.
r ( r +1) t 2
b) From Exercise S5-20, M (t ) = 1 + λr t + λ2 2! + ...
r ( r +1)
c) µ 1' = r
λ and µ 2' = λ2
which agrees with Exercise S5-20.

Section 5-10 on CD

S5-25. Use Chebychev's inequality with c = 4. Then, P( X − 10 > 4) ≤ 161 .

S5-26. E(X) = 5 and σ X = 2.887 . Then, P( X − 5 > 2σ X ) ≤ 1

4
.

The actual probability is P( X − 5 > 2σ X ) = P( X − 5 > 5.77) = 0 .

5-56
S5-27. E(X) = 20 and V(X) = 400. Then, P( X − 20 > 2σ ) ≤ 1
4
and P( X − 20 > 3σ ) ≤ 1
9
The actual
probabilities are
P ( X − 20 > 2σ ) = 1 − P ( X − 20 < 40)
60 60
= 1 − 0.05e − 0.05 x dx = 1 − − e −0.05 x = 0.0498
0 0

P ( X − 20 > 3σ ) = 1 − P ( X − 20 < 60)

80 80
= 1 − 0.05e − 0.05 x dx = 1 − − e −0.05 x = 0.0183
0 0

S5-28. E(X) = 4 and σ X = 2

1 1
P ( X − 4 ≥ 4) ≤ 4
and P( X − 4 ≥ 6) ≤ 9
. The actual probabilities are
7
e−2 2x
P ( X − 4 ≥ 4) = 1 − P ( X − 4 < 4) = 1 − x!
= 1 − 0.8636 = 0.1364
x =1
9
e− 2 2 x
P ( X − 4 ≥ 6) = 1 − P ( X − 4 < 6) = 1 − x! = 0.000046
x =1

S5-29. Let X denote the average of 500 diameters. Then, σX = 0.01

500
= 4.47 x10 −4 .
a) P ( X − µ ≥ 4σ X ) ≤ 161 and P ( X − µ < 0.0018) ≥ 15
16 . Therefore, the bound is
0.0018.

X −µ
If P ( X − µ < x) = 15
16
, then P ( σ− x < σX < σx ) = 0.9375. Then,
X X

P( −x
4.47 x10 − −4
<Z< x
4.47 x10 − 4
) = 0.9375. and x
4.47 x10 − 4
= 1.86. Therefore, x = 8.31x10 −4 .

S5-30. a) E(Y) = P( X − µ ≥ cσ )
b) Because Y ≤ 1 , ( X − µ )2 ≥ ( X − µ )2Y
If X − µ ≥ cσ , then Y = 1 and ( X − µ ) 2 Y ≥ c 2σ 2 Y
If X − µ < cσ , then Y = 0 and ( X − µ )2 Y = c 2σ 2Y .
c) Because ( X − µ ) 2 ≥ c 2σ 2Y , E[( X − µ ) 2 ] ≥ c 2σ 2 E (Y ) .
d) From part a., E(Y) = P( X − µ ≥ cσ ) . From part c., σ 2 ≥ c 2σ 2 P( X − µ ≥ cσ ) . Therefore,
1
c2
≥ P ( X − µ ≥ cσ ) .

5-57