Hi Gang,

Here's the important stuff in my script:

3dRegAna -rows 63 -cols 5 \
-xydata 5.66154 1 0 0 12.2829 ./5406_Z+orig \
-xydata 8.66154 0 0 0 -0.5771 ./5407_Z+orig \
-xydata -2.33846 0 0 0 4.3129 ./5408_Z+orig \
......
-model 5 : 1 2 3 0 -bucket 0 reg

I did center variable A, even though it is a change score (col 5). Age (col 1) was also mean-centered and the other covariates are categorical 0/1.

So in my output bucket, I'd want to look the coef #5 t-stat for the significance of my explanatory variable A. But if I'm getting a significant cluster in coef #2 t-stat and not in coef#5 t-stat, then that is just the contribution of the col 2 covariate, correct?

I'm not familiar with 3dMVM. How would I use 3dttest++ when my explanatory variable is continuous?

Thank you for all your help, Gang!

Thomas