Bob,
Thanks for reply. A couple things I forgot to clarify.
The output dataset of the dual despiking-followed-by-bandpass script we use does not have modified values in the ballpark of the raw data. Rather, it features small positive and negative integer values for each voxel. I suspect, then, that this is the result of the bandpass filtering routine (highpass and lowpass). Is *this* final product detrended by virtue of the BANDPASS filtering? In AFNI it looks like variation centered on a flat overall baseline.
Also, the -1zscore *IS* used in the creation of the brick of F-statistics. I was just wondering how appropriate that procedure is considering that the brik is being made from a set of bandpass-filtered brains and not raw data (that has variations in baseline signal from scan session to scan session).
