Dear experts,
this may be a rather basic question, but I'm running out of ideas. I'm currently analyzing data from a block design experiment containing heat stimulation; the preprocessing was done with spm5, first and second level analyses with spm8. I've run several second level tests, but the results are pretty consistantly the opposite from what I expect them to be (meaning there's always a deactivation where one would expect an activation). I've checked everything I could think of - first the contrasts, obviously, then all the other steps from import to DICOM format to second level design. Checked it, double-checked it and asked a colleague. It all seems to be correct.
Now I can't rule out that I've missed something, but it would help to know where to look. Can any of you think of a parameter anywhere in the procedure that might cause this particular mistake?
Thanks,
Iris Klossika
Central Institute of Mental Health
Mannheim, Germany
|