I am looking for expressions for the mean and variance of a random variable ($Y$) with continuous uniform distribution where the maximum depends on the product of two independent random variables, one ($X_1$) with Bernoulli distribution and the other ($X_2$) with continuous uniform distribution. The minima of both uniform distributions are zero. I.e.:

$X_1 \sim B(1, p_3/(p_2+p_3)) $

$X_2 \sim U(0,m) $

$Y \sim U(0, m-X_1X_2) $

My initial approach to the problem was to define the PDFs for $X_1X_2$ (for $0\leq X_1X_2 \leq m $) and for $Y|X_1X_2$ (for $0\leq Y \leq m-X_1X_2 $), and then, using the law of total expectation, to integrate and solve between $0$ and $m$ to find the expression for the mean. I was anticipating using the law of total variance and integrating to find the expression for the variance. However, the PDF for the product of a continuous uniform distribution and Bernoulli distribution is undefined, so this must be the wrong approach.

Can anyone help me to find expressions for the mean and variance of $Y$? Any help would be much appreciated.


1 Answer 1


You don't need to find the distribution of $Y$ to find the mean and variance.

Use the Law of Iterated Expectation (aka Law of Total Expectation, or the Tower Property).   The independence of $X_1, X_2$ shall also be invoked.

$$\begin{align}\mathsf E(Y)~&=~\mathsf E(\mathsf E(Y\mid X_1 X_2))\\[1ex] &=~ \mathsf E(\tfrac 12(m-X_1X_2))\\[1ex] &= \tfrac 12(m-\mathsf E(X_1)\mathsf E(X_2))\\[1ex] &=\phantom{\tfrac {m\,p_2}{2\,(p_2+p_3)}}\end{align}$$

For the variance, use the Law of Iterated Variance:

$$\mathsf {Var}(Y)=\mathsf E(\mathsf{Var}(Y\mid X_1X_2))+\mathsf {Var}(\mathsf E(Y\mid X_1X_2))$$

  • $\begingroup$ So, does this mean $Var(Y)=E(\frac{1}{12}(m-X_1X_2)^2) + Var(\frac{1}{2}(m-X_1X_2))$? I am stuck on the next step. $\endgroup$
    – NickM
    Sep 22, 2017 at 14:47
  • $\begingroup$ Use the rules, again. Expand the quadratic term in the expectation, apply Linearity of Expectation, and independence. Recall $\mathsf {Var}(a+bZ)= b^2\mathsf {Var}(Z)$ and apply the Law of Iterated Variance. $\endgroup$ Sep 22, 2017 at 15:11

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.