I am interested in running a mediation analysis to test whether the association between variable A (continuous behavioral variable) and whole-brain functional connectivity of seed region B is mediated by continuous behavioral variable C while controlling for factors such as movement, age, and gender.
Previous posts on the discussion board have suggested that 3dttest++ can handle this analysis.
To accomplish this end, could you please confirm that 3dttest++ should be run with each subject's z-scored whole-brain connectivity map with seed region B as inputs and Variable A and Variable C entered as covariates?
If so, could you please clarify how to interpret the output?
To get the main effects of each behavioral factor (A & C) independently, would a separate 3dttest++ need to be run without the other variable (A or C)?
Thank you!
S