To whom it may concern, I have figured out an indirect means of extracting the p-value from a conditional logistic regression (clr) run. I was able to extract the log-likelihood for the model, both with and without covariates. Twice that difference is the chisq for the clr, and then using pchisq I obtained the p-value.
So there's no need to look into the question, unless there is a more direct way to extract the p-value itself. Thanks again, Howard Alper ______________________________________________ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help