Post hoc in LME

Hi can someone help me with the interpretation of my results? i am a beginner.

I have a design with two groups and four different conditions. All subjects have gone through all conditions.

A 2x4 repeated measures design.

I have set up a simple LME:
Outcome ~ Group * Condition + (1|ID).

Now the main effects for both group and condition have been shown to be significant. There is no interaction.

My question - what does the main effect mean for the group? That the groups differ across all conditions or that the groups differ on average? So do I still need to do a post hoc analysis or not?

If so, how do I do this in R so that I can show where the group differences are for each of the four conditions?

So far I have tried with emeans:
emmeans(model, list(consec ~ Group* Condition), adjust = "bonferroni")

can this work?

Thanks a lot


Active Member
what does the main effect mean for the group?
i think it depends on the type of anova sum of squares. For type III it will work out to be an average across the other factor levels. R uses typeII be default? for post hoc I would say drop the interaction term maybe.