Hi Gwenaelle, Thank you very much for your help. I still have some questions regarding the design matrix for fslvbm. Since as you know I have four groups how can I add 3 different covariates to my design matrix? Thank you again. Andres Roman Urrestarazu MD, MSc IHP, BM University of Cambridge Department of Psychiatry Addenbrookes Hospital Box 189 Cambridge CB2 0QQ UK Telephone: 0044 01223 764421 Mobile: 0044 07 775614709 -----Original Message----- From: FSL - FMRIB's Software Library on behalf of Gwenaëlle DOUAUD Sent: Tue 2/23/2010 5:22 PM To: [log in to unmask] Subject: Re: [FSL] fslvbm problems header Hi Andres, using the Glm gui, you simply need to specify in the "General Linear Model" window that you have 4 evs in your design matrix (after switching from timeseries design to higher-level/non timeseries design in the "GLM setup" window). Cheers, Gwenaelle > Hi, > > > > Im working on a big data sample (n=222) doing VBM for > subjects at high risk of psychosis. I have come across some > issues with the design.mat since I have 4 different groups > (Psychosis, clinical high risk, genetic risk and controls) > comparing grey matter density between them. The Glm options > and FEAT usually involve a two group comparison so I was > wondering if any of you had any experience in multiple group > comparisons while creating the design matrix. Thank you very > much for your help. > > > > best wishes > > > > Andres Roman Urrestarazu MD, MSc IHP, BM > > > > University of Cambridge > > Department of Psychiatry > > Addenbrookes Hospital > > Box 189 > > Cambridge > > CB2 0QQ > > UK > > > >