 |
 |
Dear Franziska, Your contrast is effectively calculating the distance between two parallel lines ( relating this to Jeanette Mumford’s pages ) which is independent of where on the x-axis you make the measurement. Kind Regards Matthew -------------------------------- Dr Matthew Webster FMRIB Centre John Radcliffe Hospital University of Oxford On 4 Jun 2019, at 14:19, Franziska Mauter <[log in to unmask]<mailto:[log in to unmask]>> wrote: Dear FSL-community, I am currently analysing some DTI-data with TBSS and came across some open questions about mean centering variables and the demeaning option in the FSL randomise calculation. For one trial we made a 2-group analysis (n1=18, n2=21) with age and sex as regressors of no interest (contrast matrix: [1 -1 0 0//-1 1 0 0]) and tried out 3 different options: 1. with the -D option in FSL's randomise, 2. with mean centering age in the design matrix by hand without -D option, 3. with mean centered age plus -D option. Surprisingly all three looked exactly the same. For another set of subjects (n1=18, n2=17) we also tested randomise without demeaning option and no mean centered variables against using -D option in randomise, and again the results look complete the same. From what I read so far on the website of Jeanette Mumford (http://mumford.fmripower.org/mean_centering/) and especially in another post on that topic in this mailing list (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=FSL;d74ec31f.1301) I wonder how that comes? As far as I understood the mean centering process, for this analysis I should not demean any variables since I am only interested in the difference between the two groups and from the regressors I am only interested in the slope and the mean does not matter. Is that correct? Or should I still mean center age and even sex by hand? And is it just luck that the results look complete the same? Does it have to do with the amount of subjects in each group? Is it the fact that I have 2 groups that are similarly big? Or is there something else going wrong? Thanks in advance for your help, Franziska ######################################################################## 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