In this article, I want to recollect some facts about normally distriburted random variables. Let and
be two normally distributed random variables
and
. Then we have the following:
The random variable
is distributed as
.
This theorem can be proved in two manners.
In the first, you just compute the density function of and it turns out to be quite right.
In the second, you compute the moment generating function of two distributions and mutliply them. We also need to use the fact that two different distributions necessarily have different moment generating functions.