I want to compare cortical thickness in a group of Alzheimers patients's to healthy controls. Therefore I ran qdec setting the threshold to Min:1.3, Max:5 (assuming this would give me a significance level of p<=0,05). When I extract the cortical-thickness-data for the ROIs with "aparcstats2table" and perform two-sample-t-tests (assuming unequal variancies) on those, I can see highly significant (p<0,001) differences in many more ROIs, which don't show any significant differences on qdec. Does qdec use any corrections for multiple testing? which statistic test is qdec using? Why do the significancy level and the areas/ROIs showing significant differences differ between qdec and the extracted ROI-Data? 

Yours sincerely, Eva Wiedemann