Math 151.

Rumbos Spring 2014 1

Solutions to Assignment #18

1. Let X ∼ Normal(0, 1) and put Y = X 2 . Find the pdf for Y .

Solution: The pdf of X is given by

1 2
fX (x) = √ e−x /2 , for − ∞ < x < ∞.
2π
We compute the pdf for Y by first determining its cdf:

Pr(Y 6 y) = P (X 2 6 y) for y > 0

√ √
= Pr(− y 6 X 6 y)
√ √
= Pr(− y < X 6 y), since X is continuous.

Thus,
√ √
Pr(Y 6 y) = Pr(X 6 y) − Pr(X 6 − y)
√ √
= FX ( y) − FX (− y) for y > 0.

We then have that the cdf of Y is

√ √
FY (y) = FX ( y) − FX (− y) for y > 0,

from which we get, after differentiation with respect to y,

√ 1 √ 1
fY (y) = FX0 ( y) · √ + FX0 (− y) · √
2 y 2 y
√ 1 √ 1
= fX ( y) √ + fX (− y) √
2 y 2 y
 
1 1 −y/2 1 −y/2
= √ √ e +√ e
2 y 2π 2π
1 1
= √ · √ e−y/2 ,
2π y
for y > 0, where we have applied the Chain Rule. Hence,

1 1 −y/2
 √2π · √y e , for y > 0;



fY (y) =



0 for y 6 0.


Math 151. Rumbos Spring 2014 2

2. Let X and Y be independent Normal(0, 1) random variables. Put Z = X + Y .

Compute the pdf of Z.

Solution: Compute the mgf of Z:

ψZ (t) = E(et(X+Y ) )

= E(etX+tY )

= E(etX · etY ),

so that
ψZ (t) = E(etX )E(etY ),
since we are assuming that X and Y are independent. We then have
that
ψZ (t) = ψX (t) · ψY (t)
1 2 1 2
= e2t · e2t ,
since both X and Y have a standard normal distribution. Thus,
2
ψZ (t) = et
1 2
= e 2 (2)t .

It then follows by the uniqueness theorem for moment generating

functions that Z ∼ Normal(0, 2). Hence, the pdf of Z is
1 2
fZ (z) = √ √ e−x /4 , for − ∞ < z < ∞,
2π 2
or
1 2
fZ (z) = √ e−x /4 , for − ∞ < z < ∞.
2 π


3. Let X and Y be independent Normal(0, 1) random variables.

Compute Pr(X 2 + Y 2 < 1).
Math 151. Rumbos Spring 2014 3

Solution: Since X, Y ∼ Normal(0, 1), their pdfs are given by

1 2
fX (x) = √ e−x /2 , for x ∈ R,
2π
and
1 2
fY (y) = √ e−y /2 , for y ∈ R,
2π
respectively. Then, since we are assuming that X and Y are indepen-
dent, joint pdf of (X, Y ) is
1 −(x2 +y2 )/2
f(X,Y ) (x, y) = e , for (x, y) ∈ R2 . (1)
2π
Thus, ZZ
2 2
P (X + Y < 1) = f(X,Y ) (x, y) dxdy, (2)
x2 +y 2 <1

where the integrand is given in (1) and the integral in (2) is evaluated
over the disc of radius 1 centered around the origin in R2 .
We evaluate the integral in (2) by changing to polar coordinates to
get Z 2π Z 1
1 2
2 2
P (X + Y < 1) = e−r /2 rdrdθ
2π 0 0

Z 1
2 /2
= e−r rdr
0

h 2
i1
= −e−r /2
0

= 1 − e−1/2 ,
1
or Pr(X 2 + Y 2 < 1) = 1 − √ . 
e

4. Let X1 , X2 , X3 , . . . , Xn be independent identically distributed Normal(0, 1) ran-

dom. Define
Y = X1 + X2 + · · · + Xn .
Use moment generating functions to determine the distribution of Y .
Math 151. Rumbos Spring 2014 4

Solution: The moment generating function of Y is given by

ψY (t) = E(et(X1 +X2 +···+Xn ) )

= E(etX1 +tX2 +···tXn )

= E(etX1 · etX2 · · · etXn ),

so that
ψZ (t) = E(etX1 ) · E(etX2 ) · · · E(etXn ),
since we are assuming that X1 , X2 , . . . , Xn are independent. We then
have that
ψY (t) = ψX1 (t) · ψX2 (t) · · · ψXn (t)
1 2
= [e 2 t ]n
n 2
= e2t
since all Xi ’s have a standard normal distribution. It then follows by
the uniqueness theorem for moment generating functions that Y ∼
Normal(0, n). Hence, Y has a normal distribution with mean 0 and
variance n. 

5. Two instruments are used to measure the height, h, of a tower. The error
made by the less accurate instrument is normally distributed with mean 0 and
standard deviation 0.0056h. The error made by the more accurate instrument
is normally distributed with mean 0 and standard deviation 0.0044h.
Let X1 denote the measurement made by the first instrument and X2 the mea-
surement made by the second instrument. Assume that X1 and X2 are in-
X 1 + X2
dependent random variables, and let X = , the average of the two
2
instruments.

(a) Determine the distribution of X.

Solution: Write

X1 = h + E1 and X2 = h + E 2 , (3)

where

E1 ∼ Normal(0, σ12 ) and E2 ∼ Normal(0, σ22 ), (4)

Math 151. Rumbos Spring 2014 5

with
σ1 = 0.0056h and σ2 = 0.0044h. (5)
It follows from (3) and the definition of X that

E1 + E2
X =h+ . (6)
2
Setting
E1 + E2
E= (7)
2
we can rewrite (6) as
X = h + E. (8)
It follows from (7), (4), (5) and the independence of E1 and E2
that
E ∼ Normal(0, σ 2 ), (9)
where
1
σ 2 = [(0.0056h)2 + (0.0044h)2 ] = 0.00001268h2 ,
4
so that
σ = 0.00356h (10)
It then follows that X ∼ Normal(h, σ 2 ), where σ is given in (10).

(b) Compute the probability that their average of the two measurements is
within 0.005h of the height of the tower?

Solution: We want to compute the probability

Pr(|X − h| < 0.005h),

or, in view of (8),

Pr(|E| < 0.005h). (11)
Dividing both sides of the inequality in (11) by σ in (10), and using
the result in (9), we obtain from (11) that

Pr(|X − h| < 0.005h) =

˙ Pr(|Z| < 1.4), (12)

where Z ∼ Normal(0, 1).

Math 151. Rumbos Spring 2014 6

It follows from (12) that

Pr(|X − h| < 0.005h) =

˙ FZ (1.4) − FZ (−1.4)

=
˙ 2FZ (1.4) − 1;

so that

Pr(|X − h| < 0.005h) =

˙ 2(0.9192) − 1 = 0.8384.

Thus, the probability that the average of the two measurements is

within 0.005h of the height of the tower is about 84%.