model <- multinom(categorical_FOD_FODs ~ Condition_FODs + Language_used_FODs + Condition_FODs*Language_used_FODs, indvar_FODs) summary(model) write code for the post-hoc analysis, treating each level of the variables equally. Apply bonferrini including the interactions
m1 <- glht(model, linfct = mcp(Condition_FODs = "Tukey")) summary(m1)