 |
 |
Hi The outlier subject *probably* didn't affect the initial group-ICA much (you could always rerun that to check), and if that's the case then no the only thing you need to rerun is the final randomise without that subject (or with that extra confound EV). You probably don't need to remove a control as long as the groups are still (age, etc) matched, because in the null case (no group difference) there isn't a need to match group sizes for the group-ICA. Cheers. On 17 May 2011, at 20:24, Cornelius Werner wrote: > Hello list, > > we are analyzing a group of patients in comparison to controls in a > resting state experiment. What we are interested in is the correlation > of the ICs with behavioral measures. Looking at the data we find that > some of the differential spots of dual regression output might be > driven by the particularly slow performance of one patient (when > plotting the dr_stage2 maps as a PE equivalent against behavioral > measures). Now, I would like to see how the results would be without > that patient. Do I need to re-run all of DR? Then I would also need to > remove one of the healthy controls in order to maintain balance for > the initial tc-MELODIC, right? Or could I specify one extra EV in the > randomise design matrix with a single "1" for that subject in > particular? > Any help or advice is welcome! > Thanks as always, > Cornelius > > -- > Dr. med. Cornelius J. Werner > Department of Neurology > RWTH Aachen University > Pauwelsstr. 30 > 52074 Aachen > Germany > --------------------------------------------------------------------------- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ---------------------------------------------------------------------------