Standard Error

Formula

$SE(\hat{\mathbf{\beta }})=\sqrt{\text{diag}({\hat{\sigma }}^2({\mathbf{x}}^{\prime }\mathbf{x}{)}^{-1})}$

Where:

$$\begin{array}{rl}{\hat{\sigma }}^2& =\frac{SSE}{n-k}\\ SSE& =\sum _{i=1}^n(y_i-{\hat{y}}_i{)}^2\end{array}$$

Other forms

$$\begin{array}{rl}{\hat{\sigma }}^2& =\frac{e^{\prime }e}{n-k}\\ e& =y-X\hat{\beta }\end{array}$$