search for: ssicentr

...l probit models estimated using PQL under R. The anomaly is that the log-likelihoods decrease (or, alternatively -2logLik increases) as variables are added to the null model. This is unusual, and I'm trying to figure out how to interpret it. I've found some indication (e.g., at http://www.ssicentral.com/hlm/hlm00150.htm) that PQL estimation doesn't produce meaningful log-likelihoods, but I'm suspicious of that claim. Any comments or advice would be helpful. Thanks. ---------------------------------------------------------------------- Andrew J Perrin - http://www.unc.edu/~aperrin A...