Hey guys. I suck at stats excuse my dumbness. So if I do a 2x3 factorial anova. If get a significance interaction or significant rowxcolumn f score/ interaction, I perform a post hoc to test simple comparions using the cell means right ? Here, do I also perform a post hoc on the main effects using the marginal means if either was significant ? Or there’s no need ? If the f score for the interaction was not significant then I perform post hoc on the main effects in the case that either or both are Sig. is that correct ?