 |
 |
You need to use the flexible factorial model with any design with repeated measures (or two-sample t-test of difference images). In the flexible factorial model, you need to make sure you include the subject terms in the design matrix. The full factorial is not an appropriate model. Best Regards, Donald McLaren, PhD On Wed, Feb 3, 2016 at 7:30 AM, Jose P. Paredes < [log in to unmask]> wrote: > Hi, SPM users > > I am working with fMRI images. I have a control group (CG) and a pacient > group (PG), factor 1. I also have contrast images (cons) of a task 1 (T1) > and others of a task 2 (T2) for each subject, factor 2, and want to > contrast T1>T2 in CG versus PG. > > For the 2-level analysis, should it be the same if I perform a 2x2 full > factoral design with a contrast like (CG(T1)>CG(T2)) > (PG(T1)>PG(T2)) == > [1 -1 -1 1] in contrast manager, than a two-sample t-test between CG and PG > with cons images calculated from the difference between T1 and T2 cons > images? > > I tried it and don't get the same results, but I am not sure if I did the > steps all right. > > Thanks in advance. Best regards, > > José >