Hi Sandro, Yes, you can do that with randomise: EV1: Group A (code 1 for members of group A, 0 otherwise). EV2: Group B (code 1 for members of group B, 0 otherwise). EV3: Group C (code 1 for members of group C, 0 otherwise). EV4: Age EV5: Sex (coded as +1 or -1) Contrast: C1 (A<B<C): [-1 0 1 0 0] That's it! All the best, Anderson On 24 February 2018 at 02:33, Sandro Romanzetti <[log in to unmask]> wrote: > Hello, > > I would like to ask for a quick advice. I have three groups of subjects > (A,B,C) measured only once. I would like to test the hypothesis that > mean(A)<mean(B)<mean(C) including age and gender as covariates using > randomise. > > Can I do this with a single design matrix/contrast ? > > Any help is very welcome. > > Thank in advance >