emmeans_test with 2x3 ANOVA

[araikes]

Hi @kassambara,

I love the rstatix + emmeans + ggpubr environment and am using it in a lot of my analyses. One thing I've run into is using emmeans_test when I have a significant main effect in a 2x3 ANOVA. If I use emmeans outside of rstatix, the cell means are averaged over the levels of the other main effect. However, if I use emmeans_test, I do not have the option to average over the levels of another categorical variable. This results in inflated means and artificially lowered p-values.

Is there anyway to match the functionality of a straight emmeans call, e.g.:

my_aov <- lm(dv ~ pred1 * pred2, data = data)
emmeans(my_aov, pairwise ~ pred1)

Thanks