up

Creating Point Estimators - Method of Moments

A moment is … a measure of … the shape? of a function?

Populations have moments, Samples have moments, and these moments relate to each other.

Definitions

$K$th population moment = $\mu _ k$ = $E(X^k)$

$K$th sample moment = $m _ k$ = $\frac 1 n \sum (x_i^k)$

The First Moment - Easy Street

$\mu _ 1$ = $E(X^1)$ = Population Mean

$m_1$ = $\frac 1 n \sum (x_i)$ = $\bar x$ = Sample Mean

The Goal

Estimate $\theta _ 1 , \theta _ 2 , … \theta _ k$

The Method

Solve the System of Equations $\mu _ k = m _ k$

Finding the estimator for one parameter is easy. Finding two is harder, as it becomes a system of equations.

See the exponential example.