The sixth in the Bayesian Seminar series compared fitting generalized linear mixed models by classical likehlihood and Bayesian techniques.  The scripts for the examples are here mixed