An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20120908/056d6ed3/attachment.pl>
MixMod lsmeans question
2 messages · guillaume chaumet, Ben Bolker
1 day later
guillaume chaumet <guillaumechaumet at ...> writes:
Dear mixed models list, I have a question concerning TotalAnalysis function from MixMod package. Anova results tell me that there is no interaction effect between two factors but lsmeans table show me estimate values and p value as though there was an interaction effect. Is it a normal behavior?
I'm not sure, but it sounds as though you may be comparing hypothesis tests done on the basis of Wald tests (based on estimates of curvature of the goodness-of-fit surface) [i.e. "lsmeans" table] with those based on explicit model comparisons (*possibly* Anova(), depending on details ...) These can be different. More details would be helpful. Ben Bolker