 |
 |
Hi Charlie, Yep, this is correct. The group "A" is the fist when listed alphabetically (here, CON). The first contrast in the design.con file is [1 -1], which tests A>B, and this contrast is the one that appears with the suffix "1" in the file outputs. The second contrast has outputs ending with "2", and is [-1 1], testing A<B. All the best, Anderson On 7 November 2013 23:17, Charlie Leger <[log in to unmask]> wrote: > Tbss authorities, > > In tbss analysis, with a simple 2 group tbss design, a new directory was > made into which all FA images were placed. FA images were named in > logical order (as stipulated by the user guide): all controls were had the > prefix CON and a unique identifying string (e.g. CON_57dti_FA.nii.gz). The > the experimental group (or patients) were given the prefix PAT 2 and a > unique identifying string. > > With a simple 2 group design (as in the provided design_ttest2), the main > contrasts should be > > A >B, which means CON (controls) group > patients > B> A, which means PAT ( patients) group > controls > > I assume this translates to mean that the tbss_tfce_corrp_stat1 references > A>B (controls > patients) and that the > tbss_tfce_corrp_stat2 references B>A (patients > controls) > > Is this correct? > > Thank you for your help > > Charlie >