Hi Linda,

I am curious why you use 3dcalc to extract a beta of
interest in the first place. Generally, I would try to
use sub-brick selection in group tests. Note that there
is also an option of extracting or applying sub-brick
selection via labels, e.g.

3dbucket -prefix con1_beta0 stats+tlrc"[con1#0_Coef]"

or maybe:

3dttest++ -setA horses ...
subjXX horses/subjXX/subjXX.results/stats.subjXX+tlrc"[con1#0_Coef]" ...


But the -iresp output is different in any case. It
contains the response function estimate at each voxel,
and is the sum of each beta times that response curve.

SPMG2 has 2 response curves: a gamma-variate and its
derivative, and yields 2 beta weights. The iresp
output (at any voxel) is the sum of each beta times
its curve.

To answer your actual question, yes, you can get the
beta directly without using 3dcalc (see the examples
above). However, the -iresp output is not the way to
do that.

- rick