> I would now like to correlate behavioural performance to brain activation when
> these two response types are compared (i.e., to see what brain regions more
> highly active for RFacevB as compared to RScrambledvB correlate with
> behavioural performance). Do I need to compare RFacevB to RScrambledvB in
> each participant and do the correlation on these datasets?
What is your hypothesis? In other words, are you looking for the correlation between the contrast of RFacevB to RScrambledvB and the corresponding difference score of memory performance? If so, yes, feed in 3dRegAna with the contrast of RFacevB to RScrambledvB from each subject as input.
Also, you could help me test the new program 3dMEMA by donating your data (beta, t-statistic, behavior data, plus 3dRegAna script) if you don't mind: [
afni.nimh.nih.gov]
Gang