I have the following example:

Let $Y_1, \dots, Y_n$ be an i.i.d. $N(\mu, \sigma^2)$. Note that $\sum_{i = 1}^n (y_i - \mu)^2 = \sum_{i = 1}^n (y_i - \bar{y})^2 + n(\bar{y} - \mu)^2$.

We show that $Y$ and $\sum_{i = 1}^n (Y_i - \bar{Y})^2$ are independent.

One can show that

$$\begin{align} \text{Cov}(\bar{Y}, Y_i - \bar{Y}) &= \dfrac{1}{n^2} \text{Cov} \left( \sum_{j = 1}^n Y_j, nY_i - \sum_{j = 1}^n Y_j \right) \\ &= \dfrac{1}{n^2} \left( (n - 1)\text{Var}(Y_i) - \sum_{j = 1, j \not= i}^n \text{Var}(Y_j) \right) \\ &= \dfrac{1}{n^2} ((n - 1) \sigma^2 - (n - 1)\sigma^2) \\ &= 0 \end{align}$$

Since $(\bar{Y}, Y_i - \bar{Y})$ is normally distributed and this implies $\bar{Y}$ and $Y_i - \bar{Y}$ are independent for all $i$. So $\bar{Y}$ and $(Y_1 - \bar{Y}, \dots, Y_n - \bar{Y})$ are also independent. This implies $\bar{Y}$ and $\sum_{i = 1}^n (Y_i - \bar{Y})^2$ are independent.

How did the authors get from $\dfrac{1}{n^2} \text{Cov} \left( \sum_{j = 1}^n Y_j, nY_i - \sum_{j = 1}^n Y_j \right)$ to $\dfrac{1}{n^2} \left( (n - 1)\text{Var}(Y_i) - \sum_{j = 1, j \not= i}^n \text{Var}(Y_j) \right)$?

I would greatly appreciate it if people would please take the time to clarify this.


It comes from the distributive property of covariance: $$\begin{align}C&=\text{Cov} \left( \sum_{j = 1}^n Y_j, nY_i - \sum_{j = 1}^n Y_j \right)=\operatorname{cov}\left(\sum_{j=1}^n Y_j, nY_i\right)-\operatorname{cov}\left(\sum_{j=1}^n Y_j,\sum_{k=1}^n Y_k\right)\\&= n\sum_{j=1}^n\operatorname{cov}\left(Y_j,Y_i\right)-\sum_{j=1}^n\sum_{k=1}^n \operatorname{cov}(Y_j,Y_k)\\&=n\operatorname{cov}(Y_i,Y_i)-\sum_{j=1}^n\operatorname{cov}(Y_j,Y_j)=n\operatorname{var}(Y_i)-\sum_{j=1}^n\operatorname{var}(Y_j)\\&=n\operatorname{var}(Y_i)-\operatorname{var}(Y_i)-\sum_{j=1,j\neq i}^n\operatorname{var}(Y_j)\\&=(n-1)\operatorname{var}(Y_i)-\sum_{j=1,j\neq i}^n\operatorname{var}(Y_j)\end{align}$$

| cite | improve this answer | |
  • $\begingroup$ Thanks again! There seems to be an interesting property of covariance here that I was not familiar with: $\sum_{j = 1}^n \text{cov}(Y_j, Y_i) = \text{cov}(Y_i, Y_i)$. $\endgroup$ – The Pointer Mar 14 at 21:08
  • 1
    $\begingroup$ it doesn't directly come from properties of covariance. the covariance between $Y_i$ and $Y_j$ is $0$ if $i\neq j$ (because in the question it's said iid), so in the summation only $j=i$ remains. $\endgroup$ – gunes Mar 14 at 21:27
  • $\begingroup$ Ok, thanks for the clarification. $\endgroup$ – The Pointer Mar 14 at 21:28

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.