0

I know that $Var(X)=E[(X-E(X))^2]$

if $E(X)=0$,

then $Var(X)=E[X^2]$ do this equals to zero also ?

Math-fun
  • 9,507
JenuRudan
  • 103

2 Answers2

1

No, take for example a normal distribution, $X\sim N(0,1)$. The expectation is zero, the variance is known to be $1$, and hence $$1 = \text{Var}(X) = E[(X-E[X])^2] = E[X^2].$$

Em.
  • 15,981
  • And if i got X=-3Y+10 given E(Y)=0 and i need to calculate the colleration coefficient , i will get it in terms of E(X), so i can't get the value of the colleration ? – JenuRudan May 23 '16 at 10:07
  • $E[X] = E[-3Y+10]$. – Em. May 23 '16 at 10:17
1

No.

For every $\epsilon>0$ we have $$\epsilon^21_{\{|X|\geq\epsilon\}}\leq X^2$$ and consequently (take expectation on both sides):$$\epsilon^2P(|X|\geq\epsilon)\leq\mathbb EX^2$$

So $\mathbb EX^2=0$ implies that $P(X=0)=1$.

There are lots of situations where $\mathbb EX=0$ and $P(X=0)<1$.

drhab
  • 151,093