Hi- Thanks for your responses so far.
I have some questions about the output of 3dttest++ with a covariate.
- Say I have whole-brain single-subject betas from an amplitude-modulated regressor (say, trial-by-trial prediction error estimates.
- Say my covariate is some score from a questionnaire, like childhood trauma, or Beck Depression Inventory (one score per subject).
- Say I have no other between-subjects or within-subjects factors.
3dttest++ spits out a mean and a t-stat for both your dataset label (where the mean prediction-error-evoked beta for the group differs from zero) and your covariate (say, trauma score).
My question is: what do those means and t-stats REPRESENT? Do they represent the main effect of each factor, controlling for the other? Do they represent the main effect of each factor, NOT controlling for the other? Is it possible to define an interaction term, in this case? Is it possible to do this with 3dMVM? Or is it impossible, because it is not really multivariate?
Thanks,
Jim Waltz