As I understand a random variable represents all possible outcomes of an experiment with their associated probabilities. Why $X^2$ is understood as squaring outcomes of experiments instead of as multiplying the results of two identical experiments?

Let's X be a throw of a die. I've seen examples where expected value of $X^2$ is calculated as: $$ 1^2\dfrac{1}{6} + 2^2\dfrac{1}{6} + 3^2\dfrac{1}{6} + 4^2\dfrac{1}{6} + 5^2\dfrac{1}{6} + 6^2\dfrac{1}{6} = 15.16 $$

So here we just squaring the outcome of a single experiment. But in math square can be replaced by multiplication. Like: $X\times X$. And for multiplication of two random variable we probably need to construct a product distribution. As I understand it should be something like throwing 2 dice, multiplying the results, and then calculating the average of these values. So as both X are independent in this case expected value can be found as $E[XY]=E[X]E[Y]$. I.e. $3.5 *3.5 = 12.25$

  • 7
    $\begingroup$ X and X are not independent. $\endgroup$ – Michael M Jun 29 '14 at 9:32
  • $\begingroup$ X is an experiment of throwing a dice. Why two experiments like that are not independent? $\endgroup$ – Vasili Jun 29 '14 at 9:35
  • 4
    $\begingroup$ The random variables representing outcomes from rolling two die rolling experiments are independent, but that would not be written as $X\times X$. It would be $X_1 \times X_2$ - the product of the two random variables. $X\times X$ represents the square of one random variable, not the product of two different ones. $\endgroup$ – Glen_b Jun 29 '14 at 9:49
  • $\begingroup$ Make sense, thanks! $X_1 \times X_2 $ fills my gap in the notation. $\endgroup$ – Vasili Jun 29 '14 at 9:59

Your example correctly shows that $X^2$ and $X \cdot Y$ are not distributed equally. So where is the difference between the two settings?

A random variable is a fixed function that quantifies each possible outcome of a prespecified experiment.

In your example, there are two such experiments:

  1. The first one is "Roll the first die" quantified by $X = $ number of
    dots on the first die,
  2. while the second one is "Roll the second die (independent of the first)"
    quantified by $Y = $ number of dots on the second die.

Although the distributions of $X$ and $Y$ are the same (they take the values 1 to 6 each with probability $1/6$), $X$ and $Y$ are not equal since the first roll does not always equal the second roll.

The random variable $X^2$ on the other hand belongs only to the first experiment "Roll the first die" and is defined as $X^2 = $ squared number of dots on the first die. It differs from $Y$ in just looking twice at the same roll instead of looking at two different rolls.


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.