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.

Let $X_1, \cdots, X_n$ be $n$ random variables (not necessarily independent) such that $E[X_i] > E[X_j]$ whenever $i < j$. I am interested in obtaining lower bounds on the following probability: $P[ (X_1 > X_2) \wedge (X_1 > X_3) \wedge \cdots \wedge (X_1 > X_n) ]$.

share|improve this question
    
Identically distributed? –  Emre Jun 23 '11 at 19:12
    
@Emre, no they are not identically distributed –  Travis Service Jun 23 '11 at 19:13
3  
Note that "lowerbounding" is not a word. There is a single word for this -- minorizing -- but it is now somewhat obscure and old-fashioned. Good mathematical prose does not call attention to itself, so I have suggested an unobtrusive alternative. –  Pete L. Clark Jun 24 '11 at 1:19
add comment

1 Answer

The lower bound is $0$.

For example let $X_i = -(i^{k-1})$ with probability $\frac{i^2-1}{i^k-1}$ for some $k>2$, and $X_i=-\frac{1}{i}$. Then $E[X_i]=-i$, satisfying the condition.

If $i <j$ then both possible values of $X_i$ are between the two possible values of $X_j$. In particular, $X_1 = -1$ with probability 1. Meanwhile $\frac{i^2-1}{i^k-1}$ is a decreasing function of $k$, so as $k$ increases the more negative value of $X_i$ becomes less likely.

So $\Pr(X_1 > X_2)$ and the longer expression can be made arbitrarily small by increasing $k$.

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.