5.5Normal Random Variables 173
whereris a constant. Ifr=3, andVcan be assumed (to a very good approximation) to
be a normal random variable with mean 6 and standard deviation 1, find
(a)E[W];
(b)P{W> 120 }.
SOLUTION
(a) E[W]=E[ 3 V^2 ]
= 3 E[V^2 ]
=3(Var[V]+E^2 [V])
=3(1+36)= 111
(b) P{W> 120 }=P{ 3 V^2 > 120 }
=P{V>
√
40 }
=P{V− 6 >
√
40 − 6 }
=P{Z>.3246}
= 1 − (.3246)
=.3727 ■
Another important result is that the sum of independent normal random variables is
also a normal random variable. To see this, suppose thatXi,i=1,...,n, are independent,
withXibeing normal with meanμiand varianceσi^2. The moment generating function
of
∑n
i= 1 Xiis as follows.
E
[
exp
{
t
∑n
i= 1
Xi
}]
=E
[
etX^1 etX^2 ···etXn
]
=
∏n
i= 1
E
[
etXi
]
by independence
=
∏n
i= 1
eμit+σ
i^2 t^2 /2
=eμt+σ
(^2) t (^2) /2
where
μ=
∑n
i= 1
μi, σ^2 =
∑n
i= 1
σi^2
Therefore,
∑n
i= 1 Xihas the same moment generating function as a normal random variable
having meanμand varianceσ^2. Hence, from the one-to-one correspondence between