How to output odds ratios for multi-level independent variables in Proc Genmod?

djohn051 — Wed, 03 Apr 2019 13:40:28 GMT

PROC GENMOD DATA = TEMP;

CLASS ID AGE_GROUP(REF='0');

MODEL Y (EVENT = '1') = AGE_GROUP/dist=bin link = logit;

REPEATED SUBJECT = ID /TYPE = exch;

ESTIMATE 'AGE_GROUP EFFECT' AGE_GROUP 1 -1/EXP;

RUN;

Say I use the above syntax and want to know the odds of being in AGE_GROUP 1 relative to 0, as well as AGE_GROUP 2 relative to 0. Is there a way to compute that without producing a dummy variable?

topic How to output odds ratios for multi-level independent variables in Proc Genmod? in SAS Programming

How to output odds ratios for multi-level independent variables in Proc Genmod?