Hi Claire, the GLM must contain what different conditions etc you have that may influence data - so A and B should be modelled explicitly, not pooled. This is independent of what hypothesis (contrast) you're testing. Best, Brian On 2015-11-22 01:17, Claire Han wrote: > Dear experts, > > I have a question about the design matrix. > > I have conditions A, B, and C in a fMRI study. The only condition I > am > interested in is C... Putting in my GLM, conditions A and B > separately > (3 columns in total) vs. together (2 columns in total) - does this > affect my results, even bit? > > Thanks! > Claire