Hey,

I'm trying to form a new regressor (or subbrick) at the individual level in which I want to subtract baseline (fixation) from a condition of interest. Then I'd calculate a group mean from the new regressor.

The end goal is to do a between group (two groups) contrast on a small sample of pilot data.

I see that if I were to include the "-bout" command in 3dDeconvolve I could have subtracted the baseline coeffecient from the condition of interest using 3dcalc and specifying the subbricks. Unfortunately, I didn't include the "-bout" in my regression analysis. So is there another option to subtracting out the baseline info (maybe using 3dcalc)?
A second thought was whether it'd be reasonable to do the above at the goup level. However, what would be the statistical implications of using 3dttest at the within group level to subtract the baseline (using the -base1 function), then doing another 3dttest between groups?

Cheers
Steve