Hi Anderson, Thank you for your reply. Indeed that is my case - all observation within subject have the same degree of dependency. I would like to ask what is the right way to perform a Mixed Effects model (2 by 5 repeated measures design with 2 covariates per subject per condition) with SPM-preprocessed data? Is it possible to perform a Higher-level analysis without First-level analysis? I think I do need to specify GLM per subject, but this functionality is a part of First-level analysis that implies registration, which is not needed for my data since it has been normalized. What would be the way to proceed? Thank you in advance, Lara On 30 April 2018 at 15:44, Anderson M. Winkler <[log in to unmask]> wrote: > Hi Lara, > > FLAME can be used for within-subject effects and for interactions. It > requires compound symmetry, that is, that all observations within subject > have the same degree of dependency. > > All the best, > > Anderson > > > On 30 April 2018 at 08:10, Lara Todorova <[log in to unmask]> wrote: > >> Dear FSL users, >> >> >> >> I have a 2 by 5 repeated measures design for one group of 17 subjects. So >> I have 10 experimental conditions, for each of those I have a set of values >> per subject that I would like to include as covariates. Do I understand >> correctly that Flame 1/2 is the best way to proceed? >> >> >> Basically I want to enter contrast images of interest as dependent >> variable in the design matrix. For the covariate is it sufficient to add an >> additional column? Is it possible to set a nuisance regressor as well? >> >> >> >> s1 s2 s3 ... >> step1 step2 step3 Covariate >> Covariate_step1_subject1 1 0 0 1 0 >> 0 0.8 >> Covariate_step2_subject1 1 0 0 0 1 >> 0 1.8 >> Covariate_step3_subject1 1 0 0 0 0 >> 1 1.2 >> >> >> Also, my fMRI data set has already been realigned, coregistered, >> normalized and smoothed in SPM. Would it be possible to do Flame 1/2 at >> all? >> >> >> Thank you in advance. >> >> >> >> Best, >> >> Lara >> >> >