Hello,

A. It might be good for you to read the background on
the scaling step. From the single subject analysis
tutorial: AFNI_data6/FT_analysis/tutorial
see the t14_scale.txt step.


Part of that describes why it is not very important to
scale by the baseline rather than the mean. But if you
want to scale by the baseline, just add -bout and do so.

Note that with the current scale step, TENT betas should
already be interpreted as 100*(value-baseline)/mean, so
the only slight effect would be to scale by mean/baseline
(or 100/baseline). This depends on the other regressors,
too, which should be appropriately demeaned.

While there is no "-baseline" in afni_proc.py's scaling
computation, that done in the linear regression step.


B. Yes, that ancient HowTo#5 link should describe scaling
correctly. The t14.scale part of the FT_analysis tutorial
is only a few years old.

In any case, your TENT betas can already be viewed as PSC.


C. The iresp dataset is a TR-locked response function. If
your TENTs were specified as TR-locked to begin with, the
iresp betas will be identical.


D. Each t-stat volume implies a standard error for the
corresponding beta.


- rick