Hi Sara, I believe Anderson recommends running F-tests on both contrasts as default, which will test both directions. Permutation inference for the general linear model ( https://doi.org/10.1016/j.neuroimage.2014.01.060 ) explains F-tests and G-stats in the context of non-parametric statistics. Hope this helps, Kind Regards, Matthew -------------------------------- Dr Matthew Webster FMRIB Centre John Radcliffe Hospital University of Oxford > On 17 Sep 2019, at 23:08, [log in to unmask] <[log in to unmask]> wrote: > > Hi FSL team, > > I have used the design_ekaterina.ods example to create my design for a similar analysis. I have two groups (Treatment, waitlist) and two sessions (pre and post intervention) and I am mainly interested in the interaction effect. My question is should I only use the two contrasts defined in the example or should I also look at the t2>t1, and the interaction of t2>t1, G2>G1. Is this contrast somehow similar to the other one or not. I am a bit confused on that. > I do not want run too many analysis that decreases my analysis power but I also do not want to miss potential significant results. > Also, I was wondering if you could introduce a paper that explains the F-test and G-stats in more details. > > Thanks, > Sara > > To unsubscribe from the FSL list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1 > ######################################################################## To unsubscribe from the FSL list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1