I'm trying to test the 4way-interaction effect of the block (2 levels), learning condition (2 levels), group (2 levels) and age(4 levels)in a mixed effect model:
agmt.confir.mod2 <- lmer(RTNormalized~ Block * SeqType * Group * AgeRange + (1|ParticipantID), control=lmerControl(calc.derivs = FALSE,optCtrl=list(maxfun=100000)), REML = FALSE, data = RTacc.df)
I'm trying to test a particular hypothesis therefore, I'm coding priori contrast for each factor. Especially, I'm interested in testing the effect of age on the learning conditions in two groups. Hence, I consider to code repeated contrasts or the polynomial contrast to check if there is an expected trend for age. Whereas for the rest of the other variables I'm using sum coding to check the difference between two levels for each condition.
My primary concern is, whether it is okay to use different contrast coding factors within a lmer model.
