10.4. Fitting the Adjacent Categories Model as a Loglinear Model
We now know how to fit a logit model to a contingency table by fitting an equivalent loglinear model, but a natural question is “Why bother?” As we’ve just seen, the loglinear model is cluttered with “nuisance parameters” and is more cumbersome to specify in the MODEL statement. Even worse, the requirement of fitting the full multi-way interaction for the independent variables often leads to annoying convergence problems. Specifically, if there are any cell frequencies of 0 in the multi-way contingency table for the independent variables, the model will not converge. This problem does not arise when fitting the logit model directly.
Despite these difficulties, there are situations ...
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.
