Dear SPM experts,
Question 1:
I have 4 EVs in each subject’s 1st-level model (p1, p2, np1, np2), and I want to set up 4 PPI EVs (the seed time series is VOI_ABC.mat) and see activations of (1) p2 – p1, (2) np2 – np1, and (3) an interaction contrast [i.e., (p2-p1) – (np2-np1)]. Below is my setting and hope someone to confirm if the setting is correct or not (particularly I wonder if “left” and “right” in contrast 3 are appropriate).
P.VOI=[pwd filesep 'VOI_test.mat'];
P.contrast={'Omnibus F-test for PPI Analyses'};
P.Estimate=1;
P.extract='eig';
P.Tasks={'1' 'p1' 'p2' 'np1' 'np2'};
P.Weights=[];
P.analysis='psy';
P.method='cond';
P.CompContrasts=1;
P.Weighted=0;
P.Contrasts(1).name = ['ppi_p2-p1'];
P.Contrasts(1).STAT = ['T'];
P.Contrasts(1).left = {'p2'};
P.Contrasts(1).right = {'p1'};
P.Contrasts(1).Weighted = 0;
P.Contrasts(1).MinEvents = 1;
P.Contrasts(2).name = ['ppi_np2-np1'];
P.Contrasts(2).STAT = ['T'];
P.Contrasts(2).left = {'np2'};
P.Contrasts(2).right = {'np1'};
P.Contrasts(2).Weighted = 0;
P.Contrasts(2).MinEvents = 1;
P.Contrasts(3).name = ['ppi_(p2-p1)-(np2-np1)'];
P.Contrasts(3).STAT = ['T'];
P.Contrasts(3).left = {'p2' 'np1'};
P.Contrasts(3).right = {'p1' 'np2'};
P.Contrasts(3).Weighted = 0;
P.Contrasts(3).MinEvents = 1;
PPPI(P);
Question 2:
When extracting the time series in the seed region, we have to adjust data for a F contrast. In my 1st-level model, I have10 EVs, with the last 6 ones being motion regressors. So should I set eye(4) or eye (10) for this F contrast? (I guess it should be eye(4) because, in the practical example in spm8 manual, regressors were not included in this F contrast)
Question 3:
When extracting the time series in the seed region and after setting the F contrast, should I select the contrast of interest to extract the time series, or should I select exactly the F contrast to extract the time series? Or both make no differences?
Thanks. Mike
|