3
$\begingroup$

I ran a 4(A)x4(B)x2(C) ANOVA in SPSS. I got significant main effects of A, B and C, and significant interaction effects of AxB and BxC. I got no significant interaction effects of AxC and AxBxC.

My questions are: What do I do now? What post-hocs do I carry out? Should I make plots for AxB and BxC disregarding the third factor?

$\endgroup$
3
$\begingroup$

You're on the right track--the next thing you need to do is plot your means. An interaction effect means that, for example, A has a different effect depending on what the level of B is. So if A is, mm, size of hammer (large or small) and B is size of mouse (large or small) and your measure is end state of mouse after being hit with hammer, one possible interaction effect would be that both large and small mice are squashed after being hit with a large hammer, but only small mice are squashed after being hit with a small hammer. The best way to understand interaction effects is plot, plot, plot: A vs B, B vs C, and if there had been an AxBxC interaction, something like A vs B with separate graphs for each level of C.

Er. Sorry about the bizarre example, I have no idea where that came from...

$\endgroup$
  • $\begingroup$ That's a memorable example, I like it. Good, so after I make the plots, do I look at the pairwise comparisons for the significant interactions? Or is there some other post-hoc I should do? $\endgroup$ – Lou Morello Apr 24 '13 at 15:02
  • $\begingroup$ Depending on established norms in your specific field, you often don't need to do any post-hoc analyses on significant interactions unless there are unanswered questions; you can just report the interaction as significant (usually after stating it specifically like in my answer). Also, if my answer was helpful, please accept it so your question shows up as answered. $\endgroup$ – Krysta Apr 24 '13 at 15:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.