Hi Mona, Almost. An appropriate model for you to follow on the details page is the one-factor 4-levels repeated measures, as your effects repeat in each subject. You'll need 3 evs modelling your two factors and interaction rather than the evs modelling single 4-level factor. Your current design should be almost the same (I think), but won't quite get the degrees of freedom right for the ANOVA stats. Hope that helps, Eugene -- Centre for Functional MRI of the Brain (FMRIB) | University of Oxford John Radcliffe Hospital | Headington OX3 9DU | Oxford | UK Ph: +44 (0) 1865 222 523 | Mob: +44 (0) 7946 362 059 | Fax: +44 (0) 1865 222 717 -- On 2 September 2010 13:57, Mona Maneshi <[log in to unmask]> wrote: > I am resending this email, since I did not hear from you. > > > Hello FSL experts, > > I have a question regarding the higher level analysis using FEAT. I have a > repeated 2-factor 2-level ANOVA design. I have already read your web page on > FEAT details: "ANOVA: 2-factors 2-levels (Random effects)", but I could not > find my answer there. > Each subject in my study has 4 different runs (1 run in each of 2-level > TIME factor, and each of 2-level LEARNING factor). I am interested in the > main effect of LEARNING factor. Since I have 4 runs for each subject, I > thought that to increase the power of analysis, maybe it is better to run a > Fixed-effects analysis for each subject first (I have attached design.ppm > file), and then to combine COPE of LEARNING factor contrast (C1), which I am > interested in, for different subjects using mixed effect model (e.g. FLAME1) > with a simple averaging as EV. Also I can run the same mixed-effect model to > find the average of Time factor (C2) and Interaction (C2) between subjects. > Would you please let me know whether this method is correct? > > Thanks in advance, > Mona >