# If $X \sim N(\mu, \sigma ^2)$, show that $(X - \mu) / \sigma \sim N(0,1)$ [closed]

I don't know how to do this. Do I need to use converge in distribution? (I thought this can only been used if $n$ involves)

## closed as off-topic by saz, Did, user223391, Paramanand Singh, Claude LeiboviciJul 10 '15 at 7:47

This question appears to be off-topic. The users who voted to close gave this specific reason:

• "This question is missing context or other details: Please improve the question by providing additional context, which ideally includes your thoughts on the problem and any attempts you have made to solve it. This information helps others identify where you have difficulties and helps them write answers appropriate to your experience level." – saz, Did, Community, Paramanand Singh, Claude Leibovici
If this question can be reworded to fit the rules in the help center, please edit the question.

• What have you studied so far? Which topics/level etc ... the answer to these questions would help people here to provide you with an appropriate answer. – Math-fun Jul 9 '15 at 20:01
• @Math-fun, so far, random variable, joint distribution, expected values, now focus on distribution derived from the normal distribution. Very basic beginning to stats – ElleryL Jul 9 '15 at 20:06

## 4 Answers

Okay, so we know that $X\sim N(\mu,\sigma)$, looks like this: $$Pr(X<x) = F_X(x) = \int_{-\infty}^x\frac{1}{\sqrt{2\pi\sigma^2}}e^{-(\frac{t-\mu}{\sigma})^2}dt$$ Next observe that $$Pr(\frac{X-\mu}{\sigma}<x)=Pr(X<x\sigma+\mu)=\int_{-\infty}^{x\sigma+\mu}\frac{1}{\sqrt{2\pi\sigma^2}}e^{-(\frac{t-\mu}{\sigma})^2}dt$$ We then let $y = \frac{t-\mu}{\sigma}$, so $dy = \frac{dt}{\sigma}$ and $y(x\sigma+\mu) = \frac{x\sigma+\mu-\mu}{\sigma}=x$, while $y(-\infty)=-\infty$.

Hence $$Pr(\frac{X-\mu}{\sigma}<x)=Pr(X<x\sigma+\mu)=\int_{-\infty}^{x}\frac{1}{\sqrt{2\pi}}e^{-y^2}dy$$ Which is the distribution of $N(0,1)$, so $\frac{X-\mu}{\sigma}\sim N(0,1)$

Note that this is a very simple solution, there are much more insightful proofs that you can do, I just feel that for your level this should be sufficient. As practice try and do the same with the pdf, rather than the cdf as I did. $\ddot\smile$

• thanks you ! It just meets my level... ;) – ElleryL Jul 10 '15 at 1:21

The transformations in $\frac {X -\mu}{\sigma}$ are just translation and scaling, i.e., this is a change of coordinate from $X$ to $\frac {X-\mu}{\sigma}$ given by the function $X'=\phi(X)=\frac {X-\mu}{\sigma}$. More formally, you can use the Jacobian to study the effect of the change of variables in the density function:

You are making a (linear) change of variable in $X$, given by $X':=\frac{X-\mu}{\sigma}$, or in the notation of the link:

$X':=\phi(X)= \frac {X-\mu}{\sigma}$ , so that $\phi^{-1}=\sigma X'+\mu$, and

then $p(X')= p(\phi^{-1}(X))d\phi^{-1}(X)$.

Just write out the explicit formula for the gaussian (https://en.wikipedia.org/wiki/Normal_distribution)

and then perform the transformation. The algebra shouldn't be difficult.

• Actually, I think this just shows that the curve is transformed the way you want it do. However, I am not sure about the "~" for the random variable X. Perhaps this answer is not rigorous, but atleast from the perspective of what the distribution looks like it should be fine. – Andy Yao Jul 9 '15 at 20:25

One straightforward argument is to calculate the moment generating function of $(X - \mu) / \sigma$,

\begin{align} M_{(X - \mu) / \sigma}(t) &= M_{X - \mu}(t / \sigma) \\ &= M_{X}(t / \sigma) M_{-\mu} (t / \sigma) \\ &= e^{t \mu / \sigma + t^2 / 2} e^{ - t \mu / \sigma} \\ &= e^{t^2 / 2} , \end{align}

which is the standard normal m.g.f.