Introduction to Regression Procedures

Generalized Linear Models: The GENMOD Procedure

PROC GENMOD is a generalized linear modeling procedure that estimates parameters by maximum likelihood. It uses CLASS and MODEL statements to form the statistical model and can fit models to binary and ordinal outcomes. The GENMOD procedure does not fit generalized logit models for nominal outcomes. However, PROC GENMOD can solve generalized estimating equations (GEE) to model correlated data and can perform a Bayesian analysis.