Mean Squared Error equals Variance for Unbiased Estimator
Jump to navigation
Jump to search
Theorem
Let $T$ be an unbiased estimator for a population parameter of a population $P$.
Then the mean squared error for $T$ equals the variance of $T$.
Proof
Let $T$ be a general estimator.
From Mean Squared Error for Biased Estimator:
- $M = \var T + \paren {\map B T}^2$
where:
By definition, if $T$ is unbiased:
- $\map B T = 0$
The result follows.
$\blacksquare$
Sources
- 1998: David Nelson: The Penguin Dictionary of Mathematics (2nd ed.) ... (previous) ... (next): mean squared error
- 2008: David Nelson: The Penguin Dictionary of Mathematics (4th ed.) ... (previous) ... (next): mean squared error