Why wouldn't you want to demean? See Jeanette Mumford's web page on this. Also, you need to include an intercept term in your model. See http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#SingleGroupAveragewithAdditionalCovariate as an example. cheers, -MH On Wed, 2011-11-16 at 20:08 +0000, Rodrigo Perea wrote: > I think I understand what to do but if somebody could confirm that what I do is correct, it will be great. > > I want to create a design matrix lets where I use Var3 as a regressor with sex and age as a covariates in a group of individuals. I would do it without demeaning the age...is this correct? > > Lets say I have 5 subjects, my design.mat will look like: > Sex Age Var3 > 1 75 1634.75 > 0 75 1153.25 > 1 83 3151 > 0 73 1651 > 0 81 1503.25 > > > > and my design.con would look like: > > 0 0 1 > 0 0 -1 > > Thanks in advance, > Rodrigo