Exploring residuals for classification models
In datasets, where different observations lead to different model predictions the Pearson residuals will not be approximated by the standard-normal one.
But the index plot may still be useful to detect observations with large residuals.
