Hi Anderson, and thank you again for your help.
I'll probably have to give up the first level variance evaluation, but I don't see why I also have to give up the first level effect sizes. Going back to my first question: what would be the problem with choosing the min COPE between (A>B) and (A>C) for each subject, and then using OLS for the second level analysis on the min COPE image?
Many many thanks,
- Matan
|