That looks okay, but didn't you mention something
about regressors being nearly collinear in another
thread?
Is this the result of applying -GOFORIT, or were there
no 3dDeconvolve warnings for this subject?
Are there warnings in a 3dDeconvolve.err file?
Also, where are these large betas, well outside the
brain? The fact that you applied the blur only within
the brain will leave the outer areas nosier, and more
prone to outliers.
Would you send me your X-matrix via email? I would
like to take a closer look.
- rick