Hi Charles, Could please you show the design matrix and contrast and tell whether you're interested just in L-R differences or in the differences between, say, patients and controls and asymmetry? Thanks Anderson Am 05.04.14 00:51, schrieb csl: > Feedback on these two questions would be much appreciated. > > Having completed a L>R tbss_sym for a single group using the recommended -1 option in randomize, I then proceeded to test R>L by creating design.mat and design.con files in the glm_gui. In the glm_gui I simply changed the contrast from 1 to -1. Is this correct? > > The saved fsf files were used in the randomise statement identical to the original left_minus_right script except the 1 option was replaced with the new design.mat and design.con, where the con file was given a negative sign in the glm_gui: > > randomise -i all_FA_skeletonised_left_minus_right.nii.gz -m mean_FA_symmetrised_skeleton_mask.nii.gz –o RgL –d design.mat –t design.con --T2 > > The results appear to show a higher overall amount of FA rightward asymmetry, though leftward asymmetry is greater is certain areas. Is there an accurate method of determining the percentage of rightward and leftward asymmetry?