> The 3dMEMA will treat the sessions of the same subject as different subjects. That's not what I want.
>
> How can I tell the 3dMEMA not to model between-subject variability?
Your script looks fine. In your case, you're assessing the cross-session variability.