1
$\begingroup$

Hello I have been trying to figure out this question for a few hours and I am very stuck and don't know how to progress but I am pretty sure I am wrong and would greatly appreciate some help.

$X$~$N(\mu, \sigma^2)$

$\sigma = 2$, and we estimate the mean of of the distribution $E(X)=\mu$ using the average of n Random variables denoted $\bar{X} = \frac{1}{n}\Sigma_{i=1}^n X_i, X_i$~$N(\mu, \sigma^)$

assume that $X_1,X_2,...,X_n$ are dependant

If $\text{Corr}(Xi,Xj) =ρ\text{ for }i=1,...,n\text{ and }j=1, ...,n \text{ such that } i\neq j$, find Var$(\bar{X})$ (in terms of n and ρ):

firstly lets find the Cov$(X_i, X_j)$

$\begin{align} &\text{Corr}(X_i, X_j) = \frac{\text{Cov}(X_i, X_j)}{\sqrt{\text{Var}(X_i)\text{Var}(X_j)}}\\ \rightarrow &\text{Corr}(X_i, X_j)= \frac{\text{Cov}(X_i, X_j)}{\text{Var}(X)}\\ \rightarrow &\text{Corr}(X_i, X_j)\text{Var}(X)= \text{Cov}(X_i, X_j)\\ \rightarrow &\text{Cov}(X_i, X_j)= \text{Corr}(X_i, X_j)\text{Var}(X)\\ \rightarrow &\text{Cov}(X_i, X_j)= p\text{Var}(X) \end{align}$

now find $\begin{align} \text{Var}(\bar{X})&=\text{Var}(\frac{1}{n^2}\Sigma_{i=1}^nX_i)\\ &=\frac{1}{n^2}\Sigma_{i=1}^n\text{Var}(X_i)+2\Sigma_{1\leq i < j\leq n}Cov(X_i, X_j)\\ &=\frac{1}{n^2}\Sigma_{i=1}^n\text{Var}(X_i)+2 {n \choose 2} Cov(X_i, X_j)\\ &=\frac{1}{n^2}\Sigma_{i=1}^n\text{Var}(X_i)+2 {n \choose 2} p\text{Var}(X)\\ &=\frac{1}{n^2}n\text{Var}(X_i)+2 \frac{n(n-1)}{2} p\text{Var}(X)\\ &=\frac{\text{Var}(X_i)}{n}+ n(n-1) p\text{Var}(X)\\ \end{align} $

I am also asked to prove that if $X_1,X_2, \cdots, X_n$ are perfectly positively correlated that

$\text{Var}(\bar{X}) = \text{Var}(X_i) = \sigma^2$

but I don't see how that makes sense if my answer for $\text{Var}(\bar{X})$ is correct so I must be wrong

$\endgroup$

1 Answer 1

1
$\begingroup$

Somewhere you stopped applying $\frac{1}{n^2}$ to the right-hand side of your expressions

You should have ended up with $$\text{Var}(\bar{X}) =\frac{1}{n^2}\left(n\text{Var}(X)+2 \frac{n(n-1)}{2} \rho\text{Var}(X)\right) =\frac{1+(n-1)\rho}{n} \text{Var}({X})$$

which, as expected, is $\frac{1}{n} \text{Var}({X})$ when $\rho=0$, and is $ \text{Var}({X})$ when $\rho=1$

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .