so I guess I was trying to get a more general answer from researchers about interpreting results and what's comparable.

But for example if I used an analysis with 2x2 design and wanted to look specifically at how one of those factors (vol 4) related to scores on 2 tests, in 3dRegAna maybe I put:

3dRegAna -rows 30 \
-cols 2 \
-xydata 23 11 '../subject1+tlrc[4]' \
-xydata 28 12 '../subject2+tlrc[4]' \
-xydata 12 19 '../subject3+tlrc[4]' \
...
-model 2 1 : 0 \
-flof 0.01 \
-fcoef 1 RegAna_HIvLO_test1_f \
-fcoef 2 RegAna_HIvLO_test2_f \
-rcoef 1 RegAna_HIvLO_test1_r \
-rcoef 2 RegAna_HIvLO_test2_r \
-tcoef 1 RegAna_HIvLO_test1_t \
-tcoef 2 RegAna_HIvLO_test2_t \
-bucket 0 RegAna_HIvLO_bucket

Perhaps this could allow me to interpret correlations of scores on these 2 tests (whether positively or negatively correlated) to the difference between high and low conditions for one of the factors.

How could this type of analysis be replicated in another program such as 3dMVM (or, can it be?).

In 3dMVM, looking at the interaction between these variables I would expect to be a different analysis, but perhaps this is not the correct interpretation... so I am looking for help on interpreting the output from these programs with multiple levels of factors, to understand what is comparable and what may be unique to each program.

(along the lines of:
3dMVM -prefix MVM_HIvLOrep -jobs 4 \
-model "test1+test2" \
-wsVars "HIvLO*visaud" \
...
-gltLabel 1 HIvLO_test1 -gltCode 1 "HIvLO : 1*HI -1*LO test1 : " \
)

Thanks!!