Hi Jason
> When you take F-contrasts to the 2nd level for a group analysis, is it
> methodologically valid to perform a 1 sample T-test on a collection of
> F-contrasts?
To do 2nd level F tests you'll need to bring up con* images from each
subject, and enter them in a 2nd level F test; i.e., you can't bring
up F contrasts done on single subjects to a group analysis.
If you have more than one con* image per subject, you will probably
want to run a within-subjects ANOVA (specifying columns to account for
subject effects). Some general approaches to ANOVAs in SPM, as well
as how to set up within-subjects analyses, are discussed in this
report:
http://www.mrc-cbu.cam.ac.uk/people/rik.henson/personal/HensonPenny_ANOVA_03.pdf
Hope this helps,
Jonathan
|