Hello, I am trying to use R to carry out loglinear analysis. So far, starting from a previous post in this list (http://finzi.psych.upenn.edu/R/Rhelp02a/archive/33698.html), I have been able to use loglm() to generate likelihood ratio. I now try to find a function that generates all the models and another one to compare them. Thanks for any help. -- Nicolas Mazziotta The contents of this e-mail, including any attachments, are ...{{dropped}}