Yes that makes sense. Thanks alot.
One more question, if you don't mind. Let's say we use CSPLIN (3,12,4) and then do a contrast between 2 conditions at one lag. e.g:
+cond1[1] -cond2[1]
Will the coefficient that is printed for the glt at each voxel be the DIFFERENCE of the IRF's for each condition for this at this timepoint?
i.e. will the coefficient = IRF.cond1[1]-IRF.cond2[1]?
In the 3dDeconvolve documentation, it indicates that the coefficient will be equal to hr[1] - he[1]
(see pg. 45, example 1.4.4.3)
thanks again, nicola & mb