non significant interaction + significant simple effects

[the.magda]

Hi,

so I'm running a 2x2 ANOVA on my data and I have a situation where the interaction effect is not significant but the simple effects are...
what do I do? can I report it as significant simple effects without the significant interaction effect?

in fact, I have this situation quite often, with numerous variables, so an answer will be much appreciated...

sometimes I have both main effects and the interaction effect non-significant with significant simple effects...
sometimes one main is sig., the second isn't, interaction isn't and still - simple effects are...

thanks!

[hlsmith]

Typically you test for interaction effects, if they are not significant then you do not leave them in the final model unless you can establish a justification.