# bessel's correction

Given a sample, we want to find the population variance. Naively, we might want to find the Maximum Likelihood Estimation of the variance as \[ \frac{(x_i - \bar{x})^2}{n} \] where \(\bar{x}\) is the sample mean.

But this is actually a biased estimator.

## 1. see also

- degrees of freedom
- student's t test – the formulas involve degrees of freedom, for which a similar correction is relevant