Hello experts,
The design in brief: two cue types per Go and NoGo events. I'm interested in the contrast of a contrast [(CueA_NoGo > CueA_Go) > (CueB_NoGo > CueB_Go)]. The resulting images will be averaged across 4 runs for each subject with fixed effects modeling, and then taken up to a higher level group comparison with mixed effects.
This is all straightforward to me except for the "contrast of a contrast" part. I can design a first level GLM with CueA_NoGo > CueA_Go and CueB_NoGo > CueB_Go contrasts, but I'm having trouble thinking of a way to subtract the CueB contrast from the CueA contrast in the same first-level GLM. What is the best way to do this?
Instead of cramming all this into a single first level GLM, should I make CueA_NoGo > CueA_Go and CueB_NoGo > CueB_Go contrasts, take them up to a second level GLM where I make the desired [(CueA_NoGo > CueA_Go) > (CueB_NoGo > CueB_Go)] contrast using "inputs are 3d cope images", do my subject averaging at the third-level with these outputs, and run group comparisons at the 4th level? I presume this is equivalent to subtacting the CueB contrast from the CueA contrast with fslmaths before proceeding with subject level averaging?
Thanks!
Ahmet
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|