afni.nimh.nih.gov

Kallai,

Remember that the first sub-brick [#0] is the estimated response magnitude that is associated with the center value of the modulation variable. So the reverse in sign for a specific effect indicates that the average RT for this subject might be very different from the overal RT average for the whole group. Check the situation for all subjects. If group analysis is your ultimate goal, I would not worry too much about this for one specific subject.

Gang