Hi Stephanie,

I just wanted to add a couple of comments...

> In the same afni doc they normalize the timecourse to the min(200, a/b*100)
> we usually normalize to ((a-b)/b)*100 does this make a difference?

A constant shift must go directly into the constant baseline,
computed by 3dDeconvolve. This should have no other effect,
except that you would be looking at a timeseries centered
around zero (which might be misleading in some cases, as
negative would mean only less than average, not necessarily
less than baseline).

> shouldn't it be min(200,(a/b)*100)?

The parentheses around 'a/b' are impliled by the left-to-right
order of operation, so either way is fine.

- rick