Jackknife resampling is a method for estimation of the mean and higher order moments.

Given a sample ${x_i}$ of size $n$ for the distribution $X$, the jackknife resampling estimates the mean by leaving out each data point systematically. $n$ estimations of the mean will be obtained, with each of the estimations $x_i$

The mean of the sample is

The result is consistent with other sample mean methods. Jackknife estimates the variance of the sample

which is different from the direct estimation of the variance.

Jackknife is also used to estimate the bias of parameters.