Lecture #12 Order Statistics

Objectives of this Lecture:

After reading this unit, you should be able to
• understand the basic concept of order statistics (OS)
• understand about the application of OS

Example 12.1.
Let Y 1 : 5<Y 2 : 5<Y 3 :5 <. . .<Y 5: 5 denote the order statistics of a random sample of size 5 from a
distribution having PDF
−y
f ( y)= e ;
0 { ≥0
otherwise
Show that Z 1=Y 2 : 5 and Z 2=Y 4 : 5−Y 2 : 5 are stochastically independent.

Solution: Given, Y 1 : 5<Y 2 : 5<Y 3 :5 <. . .<Y 5: 5 denote the order statistics of a random sample of size
5 from a distribution having PDF
−y
f ( y)= e ; y ≥0
0 {otherwise
and the CDF becomes
y
F ( y)=∫ f (t) dt
0
y
−y −y
=∫ e dy=1−e ; y≥0
0

Now the joint PDF of 2nd and 4th order statistics when n=5 becomes
5! − y 2−1 −y − y 4−2−1
f 2,4 : 5 ( y 2 y 4 )= [1−e ] [1−e −1+e ] 2 4 2

( 2−1)!(4−2−1)!(5−4)!
[1−1+e− y ]5−4 e− y e− y
4 2 4

= 5![1−e− y ][e− y −e− y ]e−2 y e− y ; 0< y 2< y 4< ∞

2 2 4 4 2

Let us make a transformation

Z 1= y 2 and Z 2= y 4− y 2
After transformation, the joint PDF becomes
f 2,4 : 5 (z 1 z 2)= f 2,4 : 5 ( y 2 y 4 )|J |
= 5! ( e −4 z −e−5 z ) ( e−2 z −e−3 z ) ; 0 < Z 1 < ∞ ,0< Z 2 < ∞
1 1 2 2

Now the marginal density of Z 1 is

Order Statistics | Basic Concept | Lecture #12 Section-B

 Lectured by
STT453: Order Statistics Md. Kaderi Kibria, STAT, HSTU

∞
f 2 :5 ( z 1)=∫ f 2,4 :5 ( z 1 z 2)dz 2
0
∞
−4 z 1 −5 z1
=∫ 5! ( e −e ) ( e−2 z −e−3 z ) dz 2
2 2

0
− 4 z1 −5 z 1
= 20 ( e −e ) 0< Z 1< ∞
Again the marginal density of Z 2 is
∞
f 4 : 5( z 2 )=∫ f 2,4 : 5 (z 1 z 2 )dz 1
0
∞
−4 z 1 −5 z 1
=∫ 5! ( e −e )( e−2 z −e−3 z ) dz 2
2 2

0
−2 z2 −3 z 2
= 6( e −e ) 0< Z 2 < ∞
Since f 2,4 : 5 (z 1 z 2)= f 2 :5 ( z 1) f 4 :5 ( z 2) , so Z 1 and Z 2 are stochastically independent.

Example 12.2.
Let X be a random sample of size n with PDF of
1 1
f ( x )=
1;
0
− < x<
2
{
otherwise
2

Find the PDF of the sample range Z =X n : n− X 1 :n .

Solution: Given that X is a random sample of size n with PDF of

1 1
f ( x )= 1 ;
0 { − < x<
2 2
otherwise
and the CDF becomes
x
F ( x)= ∫ f (t) dt
−1 /2
x
= ∫ 1 dx= x+ 12 ; 1
− <x<
2
1
2
−1/ 2

We know the joint density function of smallest and largest order statistics is
n−2
f 1 , n: n ( x 1 x n )=n(n−1)[F ( x n)− F ( x1 )] f (x 1 ) f (x n)
1 1
= n(n−1)[ x n− x1 ] n−2 ; − < x<
2 2
Let us make a transformation
Z =x n− x1 and Y =x n
After Jacobian transformation, the joint density function of Z and Y becomes
n−2 1 1
f ( z y)=n(n−1) z ; 0< z <1 , z− < y <
2 2
Now the marginal density of Z becomes

Order Statistics | Basic Concept | Lecture #12 Section-B

 Lectured by
STT453: Order Statistics Md. Kaderi Kibria, STAT, HSTU

1 /2
n−2
f (z )= ∫ n(n−1) z dy
1
z−
2
n−2
= n(n−1)(1− z) z ; 0< z <1
This is the required density function of the range Z.

Example 12.3.
Let X 1 : 3< X 2 : 3< X 3 : 3 be the order statistics of a random sample of size 3 with PDF
f (x )= 1 ; { 0< x <1
0 otherwise
Let Z =( X 1: 3 + X 3 : 3)/ 2 be the mid range of the sample. Find the PDF of Z and also find its mean.
Solution: Given that

f (x )= 1 ; 0< x <1
0 { otherwise
and its CDF becomes
F ( x)=x ; 0 < x <1
First we have find the joint pdf of X 1 : 3 and X 3 : 3 , so the joint PDF of smallest and largest order
statistics is
n−2
f 1,3:3 ( x1 x 3 )=3(3−1)[ F ( x 3)−F ( x 1)] f ( x1 ) f ( x 3)
= 3×2 [ x 3−x 1 ]3−2 .1.1=6( x 3− x 1) ; 0 < x 1< x 3 <1
Let us make a transformation
Y 1+Y 3
Z= and W =Y 3−Y 1
2
After the Jacobian transformation, the joint PDF of Z and W becomes
f (z w)= f 1,3 :3 ( x 1 x 3)|J|
1 1
= 6[w ].1=6 w ; 0< w< 2 z , 0< z < ; 0< w< 2(1− z) , < z <1
2 2
Now the marginal density function of Z is
1
f (z )=
{ 12 z 2 ;
1
12−24 z +12 z 2 ; − < z <1
2
0< z <
2

<><><><><><><><> End <><><><><><><><>

Order Statistics | Basic Concept | Lecture #12 Section-B

 Lectured by
STT453: Order Statistics Md. Kaderi Kibria, STAT, HSTU

References
1. Balakrishnan, N. and A. C. Cohen (2004), Order Statistics and Inference: Estimation
Methods, Academy Press Inc., London.

2. Arnold, B.C., Balakrishnan, N. and Nagaraja, H.N. (2008), First Course in Order
Statistics, John Wiley and Sons, New York.

3. Anderson,T.W. (2003), An Introduction to Multivariate Statistical Analysis, 5th Ed.,

Wiley, N.Y.

4. Johnson, R. A. and Wichern, D. W (2002), Applied Multivariate Statistical Analysis, 5th

Ed., Prentice Hall, N.Y.

5. Izenman, A. J. (2008), Modern Multivariate Statistical Techniques, Springer, USA.

6. David, H.A. (2003), Order Statistics, 3rd Edition. Wiley Series, New Jersey, Canada.

Order Statistics | Basic Concept | Lecture #12 Section-B