hi everyone,
i'm doing a resting state functional connectivity analysis and am seeking expert advice. i'm happy with my pre-processing, but have a question about the analysis itself.

is there anything wrong with simply entering my seed timecourse as a regressor in spm at the first level of analysis, and then entering the resulting contrast images into a second level RFX t test? this has been done before (e.g. Waites et al 2006 Ann Neurol; 59:335-43), but i encountered skepticism about this approach at hbm last month. is it a valid method of measuring functional connectivity, or do i need to do the more commonly done correlation with Fisher z transform?

any help would be greatly appreciated,
nick bradfield

I just noticed your question. I was monitoring the open discussion, but not the "help" forum. Now I'm looking at both. If you still need an answer, I can write more.