[R-lang] Simple effects in mixed logit models

[Stefan Th. Gries]

Hi Roger

> I don’t have an interaction and so strictly speaking, I shouldn’t analyse simple effects, but reviewers have asked us to.
But you *do* have an interaction, it's just not significant
(p=0.364172), just like the effect of langcent. Unless I am missing
something, all together it seems you're left with the significant main
effect of primecent, higher values of which increase the probability
of the predicted level of primecent (probably its alphabetically
second level). The absolutely simplest way to look at this effect
would then be

assocplot(table(primecent, score))

It would probably be best though to also test whether you need both
two random effects or not.

HTH,
STG
--
Stefan Th. Gries
-----------------------------------------------
University of California, Santa Barbara
http://www.linguistics.ucsb.edu/faculty/stgries
-----------------------------------------------