Closed brentp closed 10 years ago
Anova-type comparisons are not possible as there exists no likelihood-like quantity for this type of robustness approach. For testing the fixed effects, one can use Wald-type tests (the same caveats as in the classical case apply).
if possible, it'd be nice to be able to compare rlmer fits with anova.