afni.nimh.nih.gov

Hi Gang,

thank you for your prompt reply.
We might try to consider only 15 common levels in a 3dANOVA3, but unfortunately for each subject levels vary from 15 to 22, that's why we would rather use a 2way-ANOVA...
Thanks,
Lorenzo


Gang Chen wrote:

> Hi Lorenzo,
>
> Do those 9 subjects bear the same 15 levels? If so, you'd
> better run 3-way ANOVA with 3dANOVA3 -type 5 so that subject
> variability can be accounted by treating Subject as a random
> factor.
>
> Gang