Hello,
This is a follow up question to a discussion on the list on improper variance components being derived incorrectly in certain PET designs.
Steve Fromm (Hi Steve!) wrote regarding a PET study that was a simple t-test, with one scan per subject. The conclusion of the discussion seemed to be that the correct variance parameters could be produced by using: replications over: group, and repeated measures = no. Or, just using the fMRI basic design t-test. Either of these would allow the groups to have different variances.
I wrote in previously, with a PET study that had two groups, with two scans per subject. The conclusion of that discussion was to use replications over: subjects and repeated measures = yes. As I understood it, this would allow for correlated variances across subjects within groups, and allow for correlations between before and after scans. See: http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind04&L=SPM&P=R137955&I=-3
However, the more recent discussion seems to contradict that, particularly in a situation where the groups are of different sizes. (see http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0502&L=SPM&P=R28551&I=-3, scrolling all the way down, there’s a message that discusses different group numbers).
So now I’m really confused. I don’t completely understand the format of SPM.Vi, so it’s hard to verify what’s being done. So, if someone could explain
1) What is the proper way to non-sphericity correct a PET study with 2 or more (unequal) groups, 2 scans each
2) The proper way to non-sphericity correct a PET study with 2 or more (unequal) groups, 1 scan each
3) The format of SPM.Vi.Vi
I would be extremely appreciative.
Thanks,
Allison