Take the 2-minute tour ×
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

So I'm trying to show that ${\rm Var}(Z) \le 2({\rm Var}(X)+{\rm Var}(Y))$ for $Z = X + Y$. This seems to be pretty easy to show given that $X$ and $Y$ are uncorrelated. But I'm running into trouble at this step: $$ {\rm Var}(Z) = {\rm Var}(X) + {\rm Var}(Y) + 2E[XY] - 2E[X]E[Y] $$ Normally you could say, $X$, $Y$ uncorrelated $\rightarrow E[XY] = E[X]E[Y]$, but when you cannot do this, I'm lost. Any tips?

share|improve this question

1 Answer 1

up vote 3 down vote accepted

I don't want to write out the full answer because this looks a lot like homework or self-study (and if it is indeed homework or self-study, please add the homework or self-study tag).

Hint: the maximum value that $\operatorname{cov}(X,Y)$ can have is $\sqrt{\operatorname{var}(X)\cdot\operatorname{var}(Y)}$ (the minimum value is $-\sqrt{\operatorname{var}(X)\cdot\operatorname{var}(Y)}$). Use this together with $$\operatorname{var}(X\pm Y) = \operatorname{var}(X)+\operatorname{var}(Y) \pm 2\operatorname{cov}(X,Y)$$ to see if you can get anywhere with this exercise.

share|improve this answer
    
Definitely is homework, appreciate the answer and leaving room for the learning! I guess I just needed a tip. Wasn't aware of the homework tag –  user2208604 Mar 9 at 19:56
    
So I get it to the point where we need to show that the 2*sqrt(var(X)*var(Y)) is always less than or equal to var(X) + var(Y), but I don't get how you can always say that's the case without some bounds –  user2208604 Mar 9 at 20:19
    
$\max \sigma_Z^2 = \sigma_X^2 + \sigma_Y^2 + 2\sigma_X\sigma_Y \leq 2(\sigma_X^2 + \sigma_Y^2)$ is true if and only if $\sigma_X^2 + \sigma_Y^2 \geq 2\sigma_X\sigma_Y$. I think there might be something in my answer that can be used to show that the latter assertion does always hold, and I will leave it to you to figure out what it might be. –  Dilip Sarwate Mar 9 at 20:48

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.