Dear FSL experts,
Thank you for providing the statistical test tools.
Right now I am using randomise on my test-retest reproducibility error
FA/MD skeleton maps as described on the homepage
(http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/TBSS/UserGuide). My data is
non-normally distributed.
However, as I have multiple sets of data (originating from different
processing pipelines), I am looking for an alternative to running the
unpaired two-sample ttest separately for each comparison.
Are there any other non-parametric tests with corrections for
multiple-comparisons that I could use on the FA/MD derived images
similar to the ones available through FEAT but instead just for DTI
derived images?
Thank you very much,
Barbara
Barbara Kreilkamp
PhD Candidate
University of Liverpool
|