 |
 |
Hi Manuela, I'm not sure I follow. Sounds you are doing multiple analyses and are not sure whether looking at some results, then looking again couldn't bias the analysis, is this right? But then I'm a bit confused on how the mask interacts with the components you've chosen... One way you can guard for sure against biases is by running all analyses (all 25 ICs, 59 subjects, however many designs you have) in a single run of PALM, and including the options "-corrmod" and "-corrcon". The resulting "mcfwep" files will be corrected given that all these different investigations were done. All the best, Anderson On Fri, 7 Jun 2019 at 02:21, Manuela <[log in to unmask]> wrote: > Hi FSL experts, > > I have run a group ICA and a dual regression to obtain the dr_stage_1 and > dr_stage_2 of 59 subjects. I am interested whether there are differences in > the language network relative to education. I have created a mask with > specific regions for the language network. Now the question is whether I > can just use the dr_stage_2 output and run randomize on them with my GLM? > > I am a bit confused because before I was investigating differences in > specific resting state networks which I have chosen from the GICA. > The question is whether I am allowed to look at all significant values > (all 25 components) that show up significantly because I run randomize with > a specific mask and different GLM? or whether I still have to choose > specific resting-state networks that I obtained from the group ICA like I > did before when comparing different genes? > > Any advice is very appreciated. > Thanks! > Kind regards, > > Manuela > > > ######################################################################## > > 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