Standardized (Pearson) residuals

\[\begin{equation} \tilde{r}_i = \frac{r_i}{\sqrt{\mbox{Var}(r_i)}} \end{equation}\]

where \(\mbox{Var}(r_i)\) is the estimated variance of the residual \(r_i\).

Model	Estimation Method
Classical linear-regression model	The design matrix.
Poisson regression	The expected value of the count.
Complicated models	A constant for all residuals.