Hi Marika, It ssems you may have mixed design.mat and design.con (?). If unsure, please attach these two files and paste the line used to invoke the dual_regression. All the best, Anderson On 11 October 2015 at 19:49, SUBSCRIBE FSL Marika <[log in to unmask]> wrote: > Hello experts! > > I recently ran Dual Regression with MELODIC with a contrast between three > patient groups and one overall F-test (3 contrasts and 1 F-test only were > specified n my glm). My t-stat for stage3 included up to 33 (i.e., > dr_stage3_ic000?_tfce_corrp_tstat33.nii.gz). I am not sure which tstat > would correspond to my f-test, and it appears that something went awry in > the process. I would appreciate any insights!! > > Thank you! > Marika >