1
$\begingroup$

Can anyone explain to me why the variance of the standard normal distribution is 1? I am trying to understand the mechanism behind standardising random variable. While I know minus the variable by the mean is like shifting the graph to make it centre at the origin, I don't know why dividing it by SD makes the variable having SD = 1 as well

$\endgroup$
  • 4
    $\begingroup$ It's the definition...if you have a normal distribution with any non-zero standard deviation you can rescale to get $\sigma =1 $ so it's really just a matter of units. $\endgroup$ – lulu Jan 12 at 11:05
  • $\begingroup$ If you view the SD as a thing that tells you how dispersed your distribution is around the mean, then you can understand why dividing your variable by a constant will divide your SD by this constant. So now just divide by the SD itself. Your SD has now become 1. $\endgroup$ – Bermudes Jan 12 at 11:17
0
$\begingroup$

The variance of standard normal distribution is $1$ by definition.

Concerning standardizing: if $X$ has a distribution with standard deviation $\sigma_X\neq0$ or equivalently with variance $\sigma_X^2$ then for every constant $c$ (also $c=\mathbb EX$) we have $\mathsf{Var}\left(\frac{X-c}{\sigma_X}\right)=1$ according to the rule:$$\mathsf{Var}(aY+b)=a^2\mathsf{Var}Y$$

Can you deduce this rule yourself?

Applying it on $Y=\frac{X-c}{\sigma_X}$ we get: $$\mathsf{Var}(\sigma_X^{-1}X+(-\sigma_X^{-1}c))=\sigma_X^{-2}\mathsf{Var}X=\sigma_X^{-2}\sigma_X^{2}=1$$

This means that we can write $X=\sigma_XU+\mu_X$ where $U:=\frac{X-\mu_X}{\sigma_X}$ has mean $0$ and variance $1$.

$\endgroup$
0
$\begingroup$

Let $X\sim N(\mu,\sigma^2)$ and $Z=\frac{X-\mu}{\sigma}$, then $Z\sim N(0,1)$, because: $$\mathbb E\left(\frac{X-\mu}{\sigma}\right)=\frac{1}{\sigma}\cdot \mathbb E(X-\mu)=\frac1{\sigma}\cdot \mathbb E(X)-\frac{\mu}{\sigma}=0;\\ \sigma^2\left(\frac{X-\mu}{\sigma}\right)=\frac{1}{\sigma^2}\cdot \sigma^2(X-\mu)=\frac1{\sigma^2}\cdot \sigma^2(X)=1.$$

$\endgroup$

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.