Thank you for the reply John, In fact, data of each group come from different scanners and the number of volumes/scanner is different (but on the all I've got the same number of subjects in gp A & B). But it's definitely a good idea to include that in the statistical design... thank you again Best, Cyril >Custom templates are only recommended for when the contrast in your image >differs from that of the templates, or if analysing a particularly unusual >patient population. Voxel-sizes shouldn't really matter too much. > >If you plan to use data from different scanners, then this fact should be >included in your design matrix. If you compare one group acquired on one >scanner, with another group, acquired on another, then this would prevent you >from getting significant results. The idea is that the analysis should not >simply detect scanner or sequence differences. >