 |
 |
Hi On 16 Apr 2010, at 18:45, Mark Shen wrote: > Thank you for your response. I had already checked and the faulty > subject was not the "best target" chosen; it was a subject that had > no artifact at all. In fact, after removing the faulty subject and > rerunning 'tbss_3' the same subject with no artifact was chosen as > "best target." So I still wonder why one faulty subject would apply > its artifact to all others in the analysis? Is this a inherent > problem while using the '-n' flag in tbss_2_reg? > > After removing the faulty subject and rerunning tbss3-5, the > resulting t-map and uncorrected p-map seem sensible and there are > effects in regions that we predicted. However, none of these > effects appear in the corrected p-map, even at extremely low (1-p=. > 3) threshold. Could this be explained by trying to remove a subject > after tbss2? No. Once that subject is removed (assuming you removed ALL files with that subject ID in the name from the FA subdirectory) the results will be the same as if that subject never existed. It sounds like you have an effect, but a statistically not very strong one. Cheers. > > Thanks for your help, > Mark > > On Apr 16, 2010, at 9:22 AM, Gwenaëlle DOUAUD wrote: > > Hi Mark, > > >> 1) Is it possible that 1 faulty subject with artifact could >> apply this artifact to all other subjects during tbss_2_reg >> and tbss_3_postreg? > > Hmmm, the only way it could (well that I can think of) is if you > were unlucky enough for your "best target" to be this "faulty" > subject. > It must have been the case since the subjects didn't show the > artifact after you've removed this subject and re-ran tbss3 (if you > want to be sure, after running tbss3, the name of the best target is > in best.msf, so this should change after removing the outlier) > >> 2) Is it possible to remove a subject (and all associated >> files) from a TBSS analysis after tbss_2_reg has already >> ran? > > I don't think this would be a problem. > >> 3) Could the troubleshooting attempt in #2 affect the >> statistical analysis and power to detect direct >> between-group differences? > > Unfortunately, I don't think so. Do your t-map/uncorrected p-map > seem sensible? > > Hope this helps, > Gwenaelle > --------------------------------------------------------------------------- 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 ---------------------------------------------------------------------------