Dear SPMers,
I've run an ANOVA in SPM5 that has resulted in main effects and an
interaction. I'm wondering if there is any justification for looking at the
interaction at a more liberal threshold than the main effects. Activations
observed for the interaction are significantly less robust than activations
observed for the main effects.
Thanks in advance for the help.
Best,
Siobhan Hoscheidt
|