I forgot about that question, sorry.

The -polort option specifies the polynomial degree
to use for modeling slow drifts in the MRI signal.
If you are unsure of which degree to use, use 'A'
(for automatic - 3dDeconvolve would decide).

Note that afni_proc.py defaults to the same degree
that 3dDeconvolve would apply via 'A'.

Using polort -1 means to not model any baseline
effect, even including a constant term. Such use
is uncommon.


Would you provide a little more information about
your stimuli and the regression step? Are subjects
watching a movie or something? Is this some sort of
correlation analysis?

- rick