Dear Experts,
The logic order of the interpretation of factorial design results is: A) if
main effects are significant and interactions are not significant -> analyze
main effects with main contrasts while B) if main effects are significant
and interactions are significant -> analyze simple effects.
The problem is: I've main effects and I've not interaction in voxel/cluster
X but beta plots and %sc plots show me strong evidence of interaction in X
infact if I analyze simple effects they are significant (FWE correction). I
dont't think that this result is a statistic artefact so: 1) Can I consider
it in a factorial design context or 2) Is it better consider it in a plain
comaprison context on the basis of the known literature?
Also you note that the same interaction is significant in another
voxel/cluster Y, which is far from X, so in this case I analyze simple
effects.
Thank you for your methodologic advice.
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.405 / Virus Database: 268.12.9/456 - Release Date: 25/09/2006
|