Question: If a distribution is equally likely to take the values 1 or 4, show that $s_{n-1} $ forms a biased estimator of sigma.
I started this question by finding that: $E(X)$ = 2.5, $E(X^2)$ = 8.5, and $Var(X)$ = 2.25. But after that I don't know how to do the proof. Do I need to assume a sample size, and then show how E($s_{n-1} $) is not equal to the standard deviation?