Are max and min of n iid r.v.s. independent?

  • Context: Graduate 
  • Thread starter Thread starter maverick280857
  • Start date Start date
  • Tags Tags
    Independent Max
Click For Summary

Discussion Overview

The discussion centers on the independence of the maximum and minimum of a sequence of independent and identically distributed (i.i.d.) random variables. Participants explore theoretical aspects and methods for verification, including the use of probability density functions (pdfs) and cumulative distribution functions (cdf).

Discussion Character

  • Exploratory, Technical explanation, Debate/contested, Mathematical reasoning

Main Points Raised

  • One participant asserts that the maximum and minimum are not independent, noting that the maximum is always greater than the minimum.
  • Another participant suggests using the factorization of the joint pdf or pmf to demonstrate the lack of independence.
  • A specific example is proposed involving the probabilities of the minimum and maximum being less than a certain value, leading to a conclusion that they cannot be independent if the factorization condition is not met.
  • A later reply recommends consulting statistical texts on order statistics to find general formulas that illustrate the relationship between the minimum and maximum, implying that they need not be independent.

Areas of Agreement / Disagreement

Participants generally agree that the maximum and minimum are not independent, but the discussion includes various methods and perspectives on how to demonstrate this, indicating some level of debate on the approach.

Contextual Notes

Participants mention the need for specific conditions under which the independence can be evaluated, such as the factorization of probabilities, but do not resolve the mathematical details involved.

maverick280857
Messages
1,774
Reaction score
5
Hi

Suppose [itex]X_{1}, \ldots, X_{n}[/itex] is a sequence of i.i.d. random variables. We define

[tex]X_{(n)} = max(X_{1}, \ldots, X_{n})[/tex]
[tex]X_{(1)} = min(X_{1}, \ldots, X_{n})[/tex]

Are [itex]X_{(n)}[/itex] and [itex]X_{(1)}[/itex] independent?

Whats the best/easiest way to verify this?

Thanks
Vivek
 
Physics news on Phys.org
They are not independent. The maximum is always larger than the minimum ...
 
Yeah, nice observation. Thanks :smile:
 
Suppose I wanted to show it using the factorization of the joint pdf or joint pmf, how would I do that?
 
You just have to find one example such that the cdf does not factorize.

Let m be the minimum, M the maximum, x some real number

What about P(m<x && M<x)

This is equal to only M being less than x ( because then m is automatically also less than x.

so P(m<x && M<x) = P(M<x)

For this to be equal to the factorized probability P(m<x)P(M<x) you need to have P(m<x)=1 for all real x ...which is not true:smile:
 
Thanks Pere :smile:
 
Look for a statistics text that discusses the distributions of order statistics and sets of order statistics. You will be able to find a general formula for the p.d.f of the [tex]\min \text{ and } \max[/tex] in terms of the marginal pdfs and joint pdf of the sample. Once you see that form, you will see that they need not be independent.
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 8 ·
Replies
8
Views
1K
  • · Replies 9 ·
Replies
9
Views
2K
  • · Replies 15 ·
Replies
15
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 35 ·
2
Replies
35
Views
5K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 10 ·
Replies
10
Views
3K