Probably not a good idea to send a question on a Friday evening. Resending because no responses received. Thanks for your help.

Dear FSL list,

I am running an analysis of DTI data from a large multi-site study. We have 194 subjects (127 in a disease group and 67 controls), and the data was acquired at 5 different sites across the U.S. (on the same scanner type using the same protocol). We did extensive calibration tests before and during the study (human and phantom), however I would like to test the actual full dataset and see if there are any between scanner/site differences there.

I'd appreciate if someone can provide the correct GLM model to do that.

Thank you,

Naama