I want to prove that, if $X$ is a real valued random variable with finite expected value, then:

$$\mathbb{E}[X]=\displaystyle \int_{0}^{\infty} \mathbf{P}(X \geq t)dt - \int_{-\infty}^{0} \mathbf{P} (X \leq t)dt.$$

We have that $$\mathbb{E}[X] = \mathbb {E} (X^{+}) - \mathbb{E} (X^{-})$$ and I know how to prove that if $Y$ is a non-negative r.v., then its expected value can be expressed as $$\mathbb{E}[Y]=\displaystyle \int_{0}^{\infty} \mathbf{P}(Y \geq t)dt.$$

I am having trouble expressing the second integral as the expectation of the negative part, $X^{-}$.

Can anyone help me with that?

  • $\begingroup$ Isn't this statement false for a Cauchy distribution: $P(x) = 1/(1+x^2)$? $\endgroup$ – eyeballfrog May 28 '17 at 14:51
  • 1
    $\begingroup$ Apply what you know to $X^+$ and to $X^-$, then use $X=X^+-X^-$ hence $E(X)=E(X^+)-E(X^-)$ with $$E(X^+)=\int_0^\infty P(X^+>x)dx=\int_0^\infty P(X>x)dx$$ and $$E(X^-)=\int_0^\infty P(X^->x)dx=\int_0^\infty P(X<-x)dx=\int_{-\infty}^0P(X<x)dx$$ $\endgroup$ – Did May 28 '17 at 15:10
  • $\begingroup$ That was very clear, thank you. $\endgroup$ – dimvolt May 28 '17 at 15:40
  • $\begingroup$ Even though this post is slightly different, I’d like to link it to the current choice of mother post. Also see the meta post for (abstract) duplicates. $\endgroup$ – Lee David Chung Lin Nov 13 '18 at 13:35

Rewrite the second integral with the change of variables $Y=-X$.


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.