In a two-way ANOVA, what exactly does a non-significant interaction mean? e.g. variables A and B both have significant main effects but there is no significant interaction effect.

Does this mean that performance on variable A is not related to performance on variable B?

  • 1
    $\begingroup$ Is the confusion over the interpretation of the interaction or of the significance test of a parameter? $\endgroup$ – AdamO May 24 '18 at 12:02

It means the joint effect of A and B is not statistically higher than the sum of both effects individually. Your response still depend on variable A and B, but the model including their joint effects are statistically not significant away from a model with only the fixed effects.

This is good for you because your model is simpler than with interactions.

| cite | improve this answer | |
  • $\begingroup$ You should also have a look at the confidence interval! $\endgroup$ – kjetil b halvorsen Apr 23 '17 at 13:56
  • $\begingroup$ @kjetilbhalvorsen Why do you think confidence interval is necessary here? $\endgroup$ – SmallChess Apr 23 '17 at 15:42
  • $\begingroup$ Well, it it is very wide it might include values that would be important if true. That is nice to know, and maybe tell you that you need more data. $\endgroup$ – kjetil b halvorsen Apr 23 '17 at 15:52

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.