 |
 |
Hi Hanneke, Thank you for your helpful reply. just a question about parameter inference to clarify... As I understand the values of the parameters of the model are arbitrary and relative to values within the same model space. That means that I can compare values within the same model within the same group, for example if for the model A and group 1 a value of an intrinsic connection x1 is 0.1 and another intrinsic connection y1 is 0.2 that means that for this group and model this intrinsic connection is twice as stronger. and if for model A group 2 one intrinsic connection x2 is 0.2 and the other y2 is 0.4 that means that for that group and model this connection is twice as stronger. but since the values are relative to each other only within the model space, I can't say that x2 is twice as big as x1 and y2 is twice as big as y1 because the values are arbitrary. Am I missing some thing? Thank you Ilana On Fri, Sep 21, 2012 at 5:41 PM, Hanneke den Ouden <[log in to unmask]>wrote: > Hi Ilana, > > With respect to model inference: When you run the same models for both > groups (A,B,C,D), you can compare whether the same model is optimal across > the 2 groups. > > With respect to parameter inference: Use between-subject classical > statistics if the optimal model is the same, e.g. model A. If between > groups you do not get the optimal model, e.g. model A is optimal in group 1 > and model B is optimal in group 2, then I would first apply Bayesian model > averaging on these 2 models, and then do the parameter inference on the > averaged models. > > Hope that helps! > > Hanneke > > > On Wed, Sep 19, 2012 at 7:47 AM, Ilana Podlipsky <[log in to unmask]>wrote: > >> Hello SPMrs! >> >> I'd like to compare the same DCM analysis between two different fMRI >> data sets of the same paradigm (2 subject groups, same scanning parameters >> and paradigm for both of them). >> How can I do this? what parameters are eligible for comparison? >> >> Thank you >> Ilana >> >> > > > > -- > Hanneke den Ouden, PhD > > Donders Institute for Brain Cognition and Behaviour > Radboud University Nijmegen > Kapittelweg 29, 6525 EN > office 1.36 > tel +31 24 36 68491 > > Center for Neural Science > New York University > 4 Washington Place, room 873 > New York NY 10003 > >