Let $X_1,...,X_n$ be an iid (independent and identically distributed) sample with mean $ \mu $ and variance $\sigma^2$.

How to show $$ (n-1)S^2 = \sum_{i=1}^n (Xi-\overline X)^2 = \sum_{i=1}^n (Xi-\mu ) ^2 - n(\mu-\overline X) ^2 $$

I found that if I treat $\mu$ equals $\overline X$ , it would do. But I don't think that's right.

  • $\begingroup$ Expand $X_i-\overline X=X_i-\mu+\mu-\overline X$. $\endgroup$ – Yves Daoust Apr 22 at 10:05
  • $\begingroup$ @YvesDaoust I have to expand the first X_i-$\overline X $as X_i-$\mu $+$\mu $-$\overline X$ and the second X_i-$\overline X$ as X_i-$\mu$-$\mu$+$\overline X $to get there. $\endgroup$ – gong.y Apr 22 at 10:29
  • $\begingroup$ What ? What do you mean by first and second ? $\endgroup$ – Yves Daoust Apr 22 at 10:30
  • $\begingroup$ @YvesDaoust (X_i-$\overline X$)^2 has two X_i-$\overline X$ . $\endgroup$ – gong.y Apr 22 at 10:32

Let $Y_i:=X_i-\mu$. We have $\overline Y=\overline X-\mu$.


$$\sum_i(Y_i-\overline Y)^2=\sum_iY_i^2-2\overline Y\sum_i Y+n\overline Y^2=\sum_iY_i^2-n\overline Y^2.$$


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.