Residuals

In linear regression, residual analysis is utilized to see if the linear fit of the model is appropriate through identifying poorly fitted values, with a large number of poorly fitted values indicating poor fit of the model. The benefit of residual analysis is instrumental in assessing model fit, and this benefit is needed in logistic regression. However, residuals are not normally distributed and require a more careful analysis. Forms of residuals specific to logistic regression are included in Statistica® and are described below.