Dear List,
I’m trying to do correlational analysis between BOLD and behavioral scores.
What I’d like to do is to see the correlation between change in BOLD from
the 1st to the 2nd session (i.e., 2nd > 1st) and correlated it with change
in behavioral scores from the 1st to the 2nd session. I have 7 subjects, and
I entered 1st and 2nd session cope images for each subject and set up the
model as in the attachments.
Under EV1 (pre-post), I entered individually demeaned behavioral scores. (I
computed the mean of the 1st (pre) and 2nd (post) session scores and
subtracted the mean from the 1st and 2nd session scores. Therefore, the
absolute values for each subject for 1st and 2nd session are the same.)
In the Contrasts & F-tests ta, I defined two contrasts as follows:
Title EV1 EV2 EV3 …
C1 1st>2nd 1 0 0 … (negative correlation)
C2 2nd>1st -1 0 0 … (positive correlation)
I ran the analysis, and it seemed to have gone fine. I checked the
activation images in FEAT report, and they seem to make sense, but I wanted
to make sure that I had set up the Stats correctly.
Has anyone done this type of analysis? Have I set up the Stats correctly?
Thank you in advance for your help!
-Yoshiko
|