Worked Examples

1. Let and be independently distributed normal random variables with parameters [ ]= ,

Var( ) = 2 and [ ] = 2 , Var( ) = .
a. Find the probability density function (PDF) of = min( , ).

1− ( ) = { > } = {min( , ) > } = { > , > }= { > } { > }

− − −2 −2 − −2
= > > = > >
√2 √2 √2

− − −2 −2
= > >
√2 √2

− −2
= 1−Φ 1−Φ
√2
− 2 −
=Φ Φ
√2

Then,
− 2 −
( )=1−Φ Φ
√2
and
1 − 2 − 1 − 2 −
( )= ( )= Φ + Φ ϕ
√2 √2 √2
for ∈ (−∞, ∞), where (∙) denotes the standard normal PDF and Φ(∙) denotes the
standard normal CDF.
 b. For all ∈ ℝ, calculate (max( , ) − min( , ) > ).

There are two cases here: is either the maximum or the minimum of these two random
variables. If > , then max( , ) − min( , ) = − ; otherwise,
max( , ) − min( , ) = − . Thus, defining the event as the difference between
max( , ) and min( , ) being greater than ,
( )= ( | > ) ( > )+ ( | > ) ( > )
= ( − > ) ( > )+ ( − > ) ( > )

− − [ − ] −( −2 ) − − [ − ] 0−( −2 )
= > >
Var( − ) √2 + Var( − ) √2 +

+[1 − ( − > )][1 − ( > )]

+ +
= > > + 1− > 1− >
√3 √3 √3 √3

+ +
= 1−Φ 1−Φ +Φ Φ
√3 √3 √3 √3

2. Suppose , , …, are independent and identically distributed (IID) continuous random

variables with cumulative distribution function (CDF) ( ). Define and as the maximum and the
minimum of these random variables, respectively.
a. Find the PDF of (also known as the n-th order statistics)

( ) = { < } = {max( , , …, )< }= { < , < ,…, < }

= { < } { < }… { < }= ( ) ( )… ( )=[ ( )]

( )= ( )= ( )[ ( )] for ∈ℝ

b. Find the PDF of (also known as the first order statistics)

1− ( ) = { > } = {max( , , …, )> }= { > , > ,…, > }

= { > } { > }… { > } = [1 − ( )][1 − ( )] … [1 − ( )]
= [1 − ( )]

( )= ( )[1 − ( )] for ∈ℝ

c. Are and independent?

No, they are not since the value can take must always be greater than the value of .
3. X and Y have joint density function
1
, ( , )= ≥ 1, ≥1

a. Compute the joint density function of = , = .

= ( , )= , = ( , )=

implies that

=√ = ℎ ( , ), = =ℎ ( , )

Thus,

( , )= =− ≠ 0 since ≥ 1, ≥1
1 −

Therefore,

1 1 1
, ( , )= , √ , ⁄ √ , ⁄ = ∙ =
2 2

for ≥ 1, ≥ .

,( , ) is defined on the shaded region given below (darker blue on the right), where the
vertical line is the -axis.
b. What are the marginal densities?

1 1 log
( )= , ( , ) = = ∙ log 1 = for ≥1
2 2
/

∞
⎧ 1 1 1 1
=− ∙ = for 0 < <1
⎪ 2 2 1/ 2
⎪ /
( )= , ( , ) =
⎨
⎪ 1 1 1 ∞ 1
⎪ =− ∙ = for ≥1
2 2 2
⎩