Dear TI
Yes, that design matrix would be valid. But you only get the interpretation that the first column is the mean connectivity if you mean-centre columns two and three, as you suggested.
All the best
Peter
-----Original Message-----
From: SPM (Statistical Parametric Mapping) <[log in to unmask]> On Behalf Of TI
Sent: 23 June 2021 16:26
To: [log in to unmask]
Subject: [SPM] Sp-DCM design matrix for PEB
I am sorry for posting this mail again.
I am trying to perform group comparison of DCM results within some network between three disorder groups based on PEB framework using the resting-state fMRI data.
Assume that there are three disorder groups, group A, group B and group C.
Then each group consists of 4, 3 and 2 patients respectively.
In such a case, I think the design matrix should be constructed as follows.
X=[ones(9,1),[1;1;1;1;-1;-1;-1;0;0],[1;1;1;1;0;0;0;-1;-1]]; (1)
1) The first column means the mean connectivity parameter across the three groups.
2) The second column means the group difference between the group A and the group B. (group A - group B).
3) The third column means the group difference between the group A and the group C. (group A - group C).
Then I am wondering whether to make the group difference regressor zero-mean as follows or not.
X = X(:,2:end)-mean(X(:,2:end)); (2)
Any comment is appreciated.
Thank you very much.
Best
TI
|