 |
 |
Hi Sihua, Use the command fdr: http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FDR All the best, Anderson On 27 March 2016 at 17:29, sihua xu <[log in to unmask]> wrote: > HI Anderson > > Thanks for your answer. but I mean how to do it when i consider FDR or > other correcction based on uncorectd p-valuse? > Thanks > > 2016-03-22 5:05 GMT-04:00 Anderson M. Winkler <[log in to unmask]>: > >> Hi Sihua, >> >> The results aren't significant after correction. There are no valid >> tricks that will transform a non-significant FWER-corrected result into an >> FWER-corrected significant one. That said, you could consider FDR, that >> uses the uncorrected p-values as the starting point. >> >> All the best, >> >> Anderson >> >> >> On 21 March 2016 at 02:28, sihua xu <[log in to unmask]> wrote: >> >>> Hi all >>> After completing TBSS analysis (preprocessing for eddy_correct, bet >>> and dtifit, tbss_1, tbss_2, tbss_3, tbss_4) and randomise, I used the >>> tbss_fill to visualise the results in fslview.but tbss_tfce_corrp_tstat* >>> files thresholded at p < 0.05 did not result in significant differences, I >>> used tbss_tfce_p_tstat* files (thresh at p < .005 and p < .001) and I found >>> significant differences with fslview. but I think this result is not >>> corrected, how can I get the corrected result based on this significant >>> report, just like we did whole brain correction in the SPM report, when the >>> FWE correction was not significant based on the voxel level. >>> Thanks >>> >> >> >