Why is my contrast sig.in a 1-way but no sig. interaction in a 2-way ANOVA

#1
Hi,

So I had a 2(analytical task, control task) by 2(rating: confidence, liking) design that I incorrectly put into a One-way ANOVA with planned contrasts, i.e. I made it 4 conditions. The results indicated that my when rating confidence, the analytical task did significantly worse than the control task at p=.001. There was no significant difference for liking between the analytical and control condition.

Now I've put it in a two-way ANOVA and my interaction is not significant. I have this same pattern for two different variations of the study. Can someone explain why this is happening please? Thank you so much for any help!
 

Dragan

Super Moderator
#2
Your first analysis is not appropriate because if makes no sense to do planned comparisons for a 2X2 factorial design - plain and simple.

Your second analysis is appropriate. If the interaction is not significant - as you state - then simply look at the F ratios associated with the two Main Effects. In my view, that is all that is necessary based on what you have provided above.