I am trying to determine whether independence of random variables changes when multiplied with other, potentially dependent variables. The question isn't in a measure-theoretic context.

In particular, I have 3 random variables, which we can call A, B, C

A is independent of both B and C, but B and C are not independent. Is it true then that:

$$\mathbb{E}[ABC] = \mathbb{E}[A]\mathbb{E}[BC]?$$

If it's true, I would appreciate if someone could either show me a proof (preferably one for someone without a strong measure-theoretic probability background) or point me to one.


  • 1
    $\begingroup$ Some friendly advise: you can talk about $A$ being independent of each of the variables $B$ and $C$ and you can also talk about $A$ being jointly independent of $B$ and $C$. These two are not equivalent. You cannot understand independence with intuition alone. I would strongly advise you to pick up measure theory as soon as possible, because independence cannot be understood without measure theory. $\endgroup$ – Kavi Rama Murthy Sep 18 '18 at 23:35

Let's assume you have discrete random variables so that we do not deal with integrals. First use the independence assumption: $$ \mathbb E(ABC)=\sum_{a,b,c} abc P(A=a,B=b,C=c)=\sum_{a,b,c} abc P(A=a)P(B=b,C=c). $$ And then use the following identity to get the result: $$ \sum_{a,b,c} abc P(A=a)P(B=b,C=c)=\sum_{a} a P(A=a)\sum_{b,c} bc P(B=b,C=c)=\mathbb E(A)\mathbb E(BC). $$

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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