Variation coefficient property

Click For Summary

Discussion Overview

The discussion revolves around the properties of the variation coefficient for a sum of random variables, specifically examining whether the relationship SD(T) / E(T) ≤ 1 holds for T defined as the sum of multiple random variables, T1, T2, ..., TN. The scope includes theoretical considerations and mathematical reasoning regarding independence and covariance of the random variables involved.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant states that for a random variable Ti, the inequality SD(Ti) / E(Ti) ≤ 1 holds and questions if it extends to T = T1 + T2 + ... + TN.
  • Another participant asks whether the Ti variables are independent, suggesting that independence may affect the validity of the proposed property.
  • Some participants argue that the property may hold regardless of independence, with one asserting that S(T) ≤ E(T) is true without independence assumptions.
  • A participant presents a mathematical approach to support the claim, discussing variances and covariances, and questioning if S(T) is less than or equal to the sum of individual standard deviations.
  • There is a query about the independence of N, the number of variables, and its relationship to the other random variables, raising concerns about the implications of independence in this context.

Areas of Agreement / Disagreement

Participants express differing views on whether the property holds under the condition of independence. Some support the idea that it applies generally, while others raise questions about the implications of independence and covariance, indicating that the discussion remains unresolved.

Contextual Notes

The discussion includes assumptions about the independence of random variables and the implications of covariance, which are not fully resolved. The mathematical steps presented rely on specific conditions that may not be universally applicable.

Ad VanderVen
Messages
169
Reaction score
13
TL;DR
Does a given property of a random variable also apply to the sum of that variable?
For a random variable Ti,

SD (Ti) / E (Ti) ≤ 1

with SD (Ti) = (Var (Ti))1/2 and E (Ti) the expectation of Ti and Var (Ti) the variance of Ti. My question now is whether the following property then also applies. For any variable T,

SD (T) / E (T) ≤ 1

where T = T1 + T2 + ... + TN and where N can be a fixed variable or random variable.
 
Physics news on Phys.org
Are the Tis independent?
 
Not necessarily.
 
I think the answer to your question is yes; S(T) ≤ E(T) whether or not the Tis are independent.
 
That would be great, but how can you prove this?
 
If Si ≤ Ei for all i, then
∑Si ≤ ∑Ei = E(T)
Now the question is, is S(T) ≤ ∑Si?
V(T) = ∑Vi + 2∑∑'Cij where Cij is the covariance of Ti and Tj
V(T) = ∑Vi + 2∑∑'rijSiSj where rij is the correlation coefficient between Ti and Tj
This is a maximum when all the r's = +1. So
V(T) ≤ ∑Si2 + 2∑∑'SiSj
V(T) ≤ (∑Si)2
S(T) ≤ ∑Si
S(T) ≤ E(T)
 
Ad VanderVen said:
where T = T1 + T2 + ... + TN and where N can be a fixed variable or random variable.

Suppose we have the case where ##N## is a random variable. Is ##N## known to be independent of the other random variables?

... and what would we mean by saying that ##N## is independent of a possibily infinite set of random variables ##{T_1, T_2, ...}## which are themselves not (necessarily) a set of independent random variables?
 

Similar threads

  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 0 ·
Replies
0
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K