Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I'm working on some statistics project and am not getting further because of some stupid prediction that doesn't want to be 0. That's why I was wondering if maybe the following holds: Suppose we have a random variable X and an unknown constant m. Is then $$E\left(X|X+m\right)=E(X)?$$ That would help a lot. I thought it might hold because m is unknown and thus knowing X+m doesn't say anything about X, so one could consider X and X+m as being independent?

Thanks a lot!

share|improve this question
    
I think you answered your own question. –  Memming Jul 1 '12 at 15:42
2  
Is $m$ a random variable? If not, $E[X|X+m] = X$, not $E[X]$. If $m$ is a random variable, then the answer depends on the distribution of $(X,m)$. –  madprob Jul 1 '12 at 16:06
    
@Memming: No. $ $ –  Did Jul 1 '12 at 20:24
    
If m is a constant and conditioning on X+m knowing both X and m would lead to E[X|X+m]=X But if you only know X+m and not X and m individually then it is not clear what E[X|X+m] would be but it would not be E{X] because X and X+m are highly dependent. –  Michael Chernick Jul 1 '12 at 21:17
    
Pity. Thank you guys! –  user34853 Jul 2 '12 at 11:10
add comment

1 Answer

one could consider $X$ and $X+m$ as being independent?

Certainly not. If $m$ is independent of $X$, then $X$ and $X+m$ are positively correlated.

If $m$ is constant, then $E(X|X+m)=X$ because we are conditioning on the $\sigma$-algebra generated by $X+m$, which is the same as $\sigma$-algebra generated by $X$. If $m$ is a random variable, then one should look into the joint distribuition of $X$ and $m$ to say anything about $E(X|X+m)$.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.