8.4. Fixed-Effects with Conditional Logit Analysis
Although GEE estimation is a wonderful addition to the GENMOD procedure, it doesn’t correct for bias resulting from omitted explanatory variables at the cluster level. To be more specific, when you have multiple observations per person—as in the PTSD example—it is possible to statistically control for all stable characteristics of persons, regardless of whether those characteristics can be measured. Needless to say, this is an extremely attractive possibility, and one that is easily implemented with the PHREG procedure. Unfortunately, as we shall see, there are significant limitations that keep this approach from being the universally preferred method.
I begin with a slight elaboration of the ...
Get Logistic Regression Using SAS®: Theory and Application now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
