Dear FSL experts,
I have the z-transformed maps of RSFC for two seeds (for any subjects) and I would like to test the correlation between the two using randomise, including the correlation per group interaction (cases vs controls).
How can I do this?
Is it right to use the maps of the first seed as the input file and the maps of the second seed as an EV?
To test for interaction, perhaps I have to build two merged files of maps (one with 0s in controls and the other with 0s in cases) and demeaned around the global mean of the two groups. Is it right?
Do I have to use the -d demean option in randomise, too?
Thank you for any help
Best,
Angela
|