I'm putting my feet back into the waters of SPM5 in an attempt to accomplish a comparison of group (norm vs. impaired) x task (3 item types). Which is most appropriate for determining the main effects of group and task and their interaction; the full or the flexible model? If the latter, I think I've figured out how to specify the two factors (group and task w/ task as non-independent), but I don't see where to specify the levels of the factors. And, how to enter these at the subject/scans/conditions point has me somewhat puzzled. I'm assuming that the three con.* images for the within-subject task factor should be entered in order of interest under scans and then the conditions should be expressed as vectors (?). So, for my first subject who is normal (1), then conditions should be coded as [1,1], [1,2], [1,3]. Similarly, the second subject is impaired (2) and his/her conditions would be coded as [2,1], [2,2], [2,3]. And, so on. Is this correct?
Also, for the main effects and interaction specification in the flexible factorial setup, are the main effect factor numbers just referencing the order of the factor entry into the analysis menu (e.g., group was my first factor entry, thus = 1). Is the interaction of these two factors as simple as specifying 1 2 in the GUI box?
Regards to all,
Jeff
------------------------------------
Jeff Browndyke
Duke University Medical Center
------------------------------------
|