2
$\begingroup$

Suppose $X$ is a random variable with mean $0$ and variance $\sigma_x^2$.

How can I calculate mean and variance of $X^2$?

I calculated the mean like this \begin{equation*} \operatorname{Var}(X) = \operatorname{E}(X^{2}) - \operatorname{E}^{2}(X) \rightarrow \operatorname{E}(X^{2}) = \sigma_{x}^2 \end{equation*} but am stuck at the variance.

$\endgroup$
  • $\begingroup$ It depends on the distribution of $X$. $\endgroup$ – Deep North Feb 9 '18 at 13:43
5
$\begingroup$

If you have only the mean and variance of $X$ as 0 and $\sigma_x^2$, then there is insufficient information to calculate the variance of $X^2$, which is

$$ E[X^4] - E[X^2]^2 = E[X^4] - \sigma_x^4. $$

  • For a normal R.V. $\sim N(0, \sigma_x^2)$, $E[X] = 0, E[X^2] = \sigma_x^2, E[X^4] = 3 \sigma_x^4$.

  • For a uniform R.V. $\sim U(-1.5 \sigma_x, 1.5 \sigma_x)$, $E[X] = 0, E[X^2] = \sigma_x^2, E[X^4] = 9 / 5 \sigma_x^4$.

  • It's easy to build distributions for which this is infinite.

$\endgroup$

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.