Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I want to prove the following theorem

Let $\Omega$ be an at most countable set. Let $I$ be a finite or a countable set. The set of discrete random variables $\left\{ X_{i}\right\} _{i\in I}$ auf $\Omega$ is mutually independent (i.e. for every $S_i\in \mathbb{R}$ the sets $\{X_i^{-1} (A_i) \}$ are independent) if and only if for every arbitrary finite subsets of random variables $X_{i_{1}},\ldots,X_{i_{n}}$ and values $(s_{1},\ldots,s_{n})\in\prod_{k=1}^{n}X_{i_{k}}(\Omega)$ we have $$ P(X_{i_{1}}=s_{1},\ X_{i_{2}}=s_{2},\ldots,\ X_{i_{n}}=s_{n})=\prod_{k=1}^{n}P(X_{i_{k}}=s_{k}). $$

One implication ($\Rightarrow$) is immediate. But on the other one I'm stuck. I managed to show $$ P\left(\left\{ X_{i_{1}}\in A_{1}\right\} \cap\ldots\cap\left\{ X_{i_{n}}\in A_{n}\right\} \right)=P\left(\bigcup_{x_1\in A_1,\ldots,x_n \in A_n}\left[X_{i_{1}}^{-1}\left(x_{1}\right)\cap\ldots\cap X_{i_{n}}^{-1}\left(x_{n}\right)\right]\right). $$

But the union on the right side does not have be a union of disjoint sets, so I can't apply $\sigma$ additivity - which then would help me to use my assumption. Could you please tell me how to proceed ?

I know that the proof of this theorem is a standard one, with proofs of it books like Resnick or Gut, but both of these use measure-theoretic machinery, of which I don't know anything, so I can't follow those proofs; I would just like to prove it in the simple discrete setting.

share|cite|improve this question
up vote 0 down vote accepted

Call $E_{x_1,\dots,x_n}$ the event $\bigcap_{j=1}^nX_{i_j}^{-1}(\{x_j\})$. We can see that $E_{x_1,\dots,x_n}\cap E_{x'_1,\dots,x'_n}=\emptyset$ whenever $(x_1,\dots,x_n)\neq (x'_1,\dots,x'_n)$, because at least one coordinate differs.

share|cite|improve this answer
Ah yes, of course - though I think you meant $E_{x_1,\dots,x_n}\cap E_{x'_1,\dots,x'_n}=\emptyset$ since $E_{x_1,\dots,x_n}\neq E_{x'_1,\dots,x'_n}$ doesn't yet allow one to use $\sigma$-additivity. – MyCatsHat Dec 30 '12 at 10:27
@user26698 You are right, I fixed it. Thanks! – Davide Giraudo Dec 30 '12 at 11:24

Your Answer


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.