Moment

Also known as: Raw moment, Population moment, Statistical moment

A moment of a random variable $X$ is an expectation of a power: the $j$-th raw moment is $\mu’_j=E[X^j]$. The first moment is the mean; the second relates to variance via $\mathrm{Var}(X)=E[X^2]-(E[X])^2$. Central moments use $E[(X-\mu)^j]$.

Method of moments matches population moments $\mu’_j(\theta)$ to sample moments $m’_j=n^{-1}\sum x_i^j$. In time series and method-of-moments estimators, “moments” means these expectations — not physical torque.

← Back to CS Glossary