Hi AFNI experts,
I am running a gPPI analysis similar to what is described here (https://afni.nimh.nih.gov/CD-CorrAna). The task is a simple n-back task, so I have separate PPI interaction regressors for 2-back and 0-back, respectively, on the first level. The study is an intervention study with an experimental arm and a control arm, so I am assessing for a group x time intervention in the group analysis; so in this case, I test for a condition x group x time interaction in a 3dLMEr model.
My question is the following, I would like to next extract beta weights from a ROI from the first level model (for each participant) for further analysis. However, given this gPPI approach, there is not a single beta-weight here that captures my comparison of interest (2-back vs. 0-back). What I thought of doing as a proxy for this is to extract the 2-back beta weights, extract the 0-back beta weights, and then subtract them; and do further analysis on this value. However, before going further, I wanted to see if this was a reasonable approach, or if you had any alternative recommendations.