sdm-help-list > one tailed or two tailed test
Showing 1-3 of 3 posts
Display:
Results per page:
Jul 10, 2021  03:07 PM | chen chengfeng
one tailed or two tailed test
Hello!

I have a simple question: Does AES-SDM, SDM-PSI, and GingerALE use one tailed or two tailed test?
Aug 10, 2021  09:08 AM | Janik Goltermann
RE: one tailed or two tailed test
I can't say anything to AES-SDM or ALE but in SDM-PSI one-tailed tests in both directions are conducted (the original and the inverse contrast). I double checked that using t-maps of the original and the inverse one-tailed contrasts from SPM which produced the same results in SDM-PSI (just showing the original contrast on the top of the browser result window and the inverse contrast of the original map on the bottom of the browser results). That also seems to be the reason why a p-threshold of 0.025 is the default when thresholding your results for FWE-corrected p-values (2x 1-tailed tests with an alpha of .05 each). 
Best regards
Sep 20, 2021  02:09 PM | chen chengfeng
RE: one tailed or two tailed test
It is appreciated for answering my question in your busy time. "This procedure theoretically only tests hypothesis in one direction(e.g., patients>controls), but not the hypothesis in the other direction(e.g., patients < controls ) and thus, the procedure should be conducted twice, one for each direction" (Albajes-Eizagirre, A.2019). ALE analyses is also computed with the text file(e.g., patients>controls) and the other text file(e.g., patients < controls). Consequently, for my part, ALE might also conduct the one tail text. Again, thanks for taking your valuable time to answer my questions.