Visualizing Main and Interaction Effects

Note:
  • Make sure that you comment about each type of effect (interaction, row factor, and column factor) even if that effect is not evident. In other words, explicitly state that there is no effect if there is no effect; don’t just leave that effect absent from your answer.
  • If there is an interaction effect, do NOT say that there is “no main effects.” Rather say that the “main effects cannot be assessed because of the interaction effect.”
  • The “lines” do not have to intersect to represent an interaction, they simply do not “track together.” Imagine “lines” that start near the same point but one goes up and the other goes down. They don’t intersect but they certainly don’t track together.
  • Pay close attention to my descriptions below. Come see me if they do not make sense to you.
  1. There is NO interaction effect because all “lines” “track together.” There is a row factor main effect because all “lines” in the left plot are not horizontal (implying a difference in row level means). There is a column factor main effect because all “lines” in the right plot are not horizontal (implying a difference in column level means).
  2. There is NO interaction effect because all “lines” “track together.” There is a row factor main effect because all “lines” in the left plot are not horizontal (implying a difference in row level means). There is NO column factor main effect because all “lines” in the right plot are horizontal (implying that the column level means are all equal).
  3. There IS an interaction effect because the “lines” do not “track together” (in fact they intersect). I will not comment on the main effects because of the presence of the interaction effect.