0
$\begingroup$

I can't find anywhere what the means of sums rules are so i'm confused with this question

The random variables $X_1......X_5$ are jointly multivariate normal. Their expectations are $E(x)= \mu_i$ and variances are $\sigma_i^2$ for $i=1,2....5$ respectively. The correlations are: $Corr(X_i,X_j)= \frac {E(X_iX_j)-\mu_i\mu_j}{\sigma_i\sigma_j} = \rho$ for all $1 \le i\ne j \le 5$.

a) Derive the mean of $\bar X = \sum_{1 \le i \le 5} \frac {X_i}{n}$ using all the rules for computing the means of sums

b) Derive the variance of $\bar X$ of part a) using all the rules for computing variances of sums

$\endgroup$
1
$\begingroup$

Let $X_1,..., X_n$ be random variables, not necessarily independent.

Then

$$E(\sum_{i=1}^n X_i) = \sum_{i=1}^n E(X_i)$$

$$var(\sum_{i=1}^n X_i) = \sum_{i=1}^n var(X_i) + 2\sum_{i \neq j}cov(X_i, X_j)$$ $$= \sum_{i=1}^n var(X_i) +2\sum_{i \neq j}\sigma_i\sigma_j corr(X_i, X_j).$$

EDIT So in this case

$$var(\overline{X}) = \frac{1}{n^2}var(\sum_{i=1}^n X_i) = \frac{1}{n^2} (\sum_{i=1}^n\sigma_i^2 + 2\rho \sum_{i \neq j}\sigma_i\sigma_j )$$

$\endgroup$
  • $\begingroup$ So I just did the standard $E(\bar x)=E(\frac {X_1...X_5}{n})$ then $\frac {1}{n} E(X_1)+E(X_2)+...E(X_5)$ then $\frac {1}{n} * n\mu = \mu$ is this correct or is it different because these are multivariate $X_i$ $\endgroup$ – moku Mar 25 '14 at 19:23
  • $\begingroup$ Yes, that's exactly right (if you put brackets around the sum $E(x_1 +...+X_n)$) $\endgroup$ – Frank Mar 25 '14 at 19:28
  • $\begingroup$ alright thxs mate. So there is no difference even tho the $X_i$'s represent distributions and not observations from one sample $\endgroup$ – moku Mar 25 '14 at 19:30
  • $\begingroup$ I'm not sure what you mean, but for any collection of RV's $X_1,...,X_n$, whether they have the same distribution or not, or are independent or not, the above formulas hold. $\endgroup$ – Frank Mar 25 '14 at 19:32
  • $\begingroup$ so what is the $corr(X_i,X_j)$ simplified further like $Var(X_i)= \frac {\sigma^2}{n}$ $\endgroup$ – moku Mar 25 '14 at 19:50

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.