I am supposed to prove that

Let $X \ge 0$ be a random variable defined on $(\Omega, \mathcal{A}, > P)$ and $\mathbb{E}[X] = 1$. Define $Q: \mathcal{A} \to \mathbb{R}$ by $Q(A) = \mathbb{E}[X1_{A_n}]$. Show that Q is a probability measure on $(\Omega, \mathcal{A})$

Of course $Q(\Omega) = 1$. Now let $A_n \in \mathcal{A}$ pairwise disjoint events. Then

$$Q\left(\bigcup A_n\right) = \mathbb{E}\left[X1_{\bigcup A_n}\right] = \mathbb{E}\left[\sum X1_{A_n}\right].$$

Now can I say that $\mathbb{E}[\sum X1_{A_n}] = \sum \mathbb{E}[X1_{A_n}] = \sum Q(A_n)$, completing the proof?

I know that expectation is a linear operator, but how does it behave with countable sum?

  • $\begingroup$ Try to prove the two inequalities instead.. $\endgroup$ – Snufsan May 17 '14 at 17:06

You can pass to countable sums applying theorems like Monotone Convergence or Dominated Convergence. In your case, the monotone convergence theorems does the job, once the functions defined by

$$f_n = \sum_{i=1}^n X1_{A_i}$$

converge to

$$f = \sum_{i=1}^{\infty} X1_{A_i}$$

monotonically, right?

Then you have

$$\mathbb{E}[f] = \lim_n \mathbb{E}[f_n] = \lim_n \sum_{i=1}^n \mathbb{E}[X1_{A_i}] = \sum_{i=1}^{\infty} \mathbb{E}[X1_{A_i}]$$

  • $\begingroup$ Right. Though one should also take the functions to be positive, right? $\endgroup$ – Ant Jun 1 '14 at 18:16
  • 1
    $\begingroup$ Right! You should have the right hypothesis to apply the monotone convergence theorem. In this case, $X$ Positive. $\endgroup$ – Rodrigo Ribeiro Jun 1 '14 at 21:40

Expectation is given by the integral, i.e. for a random variable $X$ on a probability space $(Y,M,P)$ we have

$\mathbb{E}X = \int_Y X dP $

Because of $X \geq 0$, you can now apply the monotone convergence theorem to complete your proof.


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.