Can someone give an example of two r.v's $X,Y$ such that $E(XY)\neq E(X)E(Y)$? I know that this means $Cov(X,Y)\neq 0$, so I've been trying r.v's where we have $Y=aX+b$ for some constants $a,b$, but I keep getting $E(XY)=E(X)E(Y)$ for these examples.
Asked
Active
Viewed 16 times
0
-
1Can you show your calculation? $\mathrm{Cov}(X,Y) = a^2\mathrm{Var X}$ for your example. – Andrew Aug 16 '23 at 18:01
1 Answers
0
Consider: $(X, Y) = (1, 0)$ with $1/2$ probability, and $(0, 1)$ with $1/2$ probability. Then $E[X] = 1/2$, $E[Y] = 1/2$, $E[XY] = 0$.
Note that in this case, $Y=1-X$.
isaacg
- 937