Thanks, Alfonso! You're the best.
Just a bit of clarification:
1. Are these instructions for a seed-voxel analysis (since you mention clusters)? Since I am running an ROI-ROI analysis, I input the contrast [0 0 1] and then "import values as covariate" rather than "extract values" (which I don't see as an option) and then look at the [-1 1 0] and [0 0 1] contrasts in the calculator, right?
2. If the severity actually decreases with higher numbers, do I then compare it the [1 -1 0] contrast rather than the [-1 1 0] contrast? Or do I just want to see an opposite sign in that case? Also, if the sign is different, but significant, do I make the opposite conclusion (differences in connectivity between groups decreases with increasing severity)? Sorry if this is a stupid question.
3. If the sign is the same, but the [-1 1 0] contrast is not significant, then can I assume that although the covariate affects connectivity in the patients, it does not affect the difference in connectivity between groups? Again, sorry if this is a dumb question.
I hope this makes sense.
Thanks again!