Hello, By including these 3 nuisance variables in the model you are controlling for them; your contrast effectively tests the data for non-zero mean after accounting for any unique variance due to age, sex, hand. Kind Regards Matthew -------------------------------- Dr Matthew Webster FMRIB Centre John Radcliffe Hospital University of Oxford > On 9 Dec 2017, at 16:33, Tia L. <[log in to unmask]> wrote: > > Dear all, > > I would like to control for nuisance variables (age, sex, hand) at group level. I have demeaned the values for these and entered them as covariates. Whilst I understand how to find correlations/additional effect of these, what I want is to control for them - i.e. get results independent of age, sex, hand. Would I achieve this by simply entering the 4 EVs (group mean + 3 additional measures) and a single contrast: 1 0 0 0? If not, how? > > Thank you very much in advance! > > Tia.