Predictor Summary

The predictor summary table shows the predictors in rank order according to the statistics used to determine the ranking. There are a mix of scale and categorical predictors, so the ranking is based on the significance values of statistical tests. Note that the significance values are not adjusted for the number of tests performed by the procedure.

Significance values for scale predictors are based upon an F test that compares the mean value of the predictor across the levels of the response.
Significance values for categorical predictors are based upon a chi-square test of the independence of the predictor and response.

By default, only the selected predictors are displayed, and the procedure chooses a maximum of 30 predictors (this ceiling is raised when there are more than 232 possible predictors) whose significance values are less than 0.05. The largest significance value in the table is 0.037, so there may be more than 30 predictors that make the significance value cutoff.