Dear gPPI forum users,

I am using for the first time gPPI analysis and despite reading the forums and doing the Tutorials, I am still uncertain about how to define the P.Tasks.

My question is: how do I define the P.Tasks for a block design (with 3 conditions: words learnt, faces learnt, rest) for which the 1st level GLM consisted of an event-related analysis?

For each participant, the event-related SPM.mat file contains 6 conditions (Words remembered, words familiar, words forgotten, faces remembered, faces familiar, faces forgotten). The event-related script generated 12 contrasts for each participants. I am only interested in 2 contrasts (words remembered and faces remembered) for the gPPI analysis.

Based on the forum's answers, I have written all of the 6 event-related conditions from the SPM.mat in P.Tasks (Words rem, Words fam, Words fo, Faces rem, Faces fam, Faces fo), and I am extracting the mean activation of voxels.
To look at connectivity modulated by the condition 'word remembered, I put 'word remembered' on the left and 'None' on the right.

Is this correct? Or should I only be writing P.Tasks={'1' 'WRem' 'FRem'} if these are the only 2 contrasts I am interested in? I would like to assess the connectivity of the memory network for words and faces which are subsequently remembered. I am not interested in items that are familiar or forgotten. 

Any feedback/answer would be greatly appreciated it!

PPI.Estimate=1;
PPI.contrast='Omnibus F-test for PPI Analyses';
PPI.extract='mean';
PPI.Tasks={'1' 'WRem' 'WFam' 'WFo' 'FRem' 'FFam' 'FFo'}; %or {0 ' WRem' 'WFam' 'WFo' 'FRem' 'FFam' 'FFo'}
PPI.Weights=[];
PPI.analysis='psy';
PPI.method='cond';
PPI.CompContrasts=1;
PPI.Weighted=0;
PPI.Contrasts(1).left={'WRem'};
PPI.Contrasts(1).right={'None'};
PPI.Contrasts(1).STAT='T';
PPI.Contrasts(1).Weighted=0;
PPI.Contrasts(1).MinEvents=5;
PPI.Contrasts(1).name='WRem';
PPPI(PPI)

% Cannot put PPI.Contrasts(2).left={'FRem'} because different VOI


Thank you!

Marine

Originally posted by Marine Fleury:

Dear Dr McLaren, 

Thank you very much for your fast and very helpful reply!

Kind regards,
Marine