A simple question regarding to xjview and slice viewer

Submitted by lixiangcx on Mon, 11/11/2013 - 18:50

        In the xjview and the slice viewer, even if I set the same p value (and the same cluster size), the thresholds are different. Why?

       For example:

       

       In xjview, if I set p = 0.01, cluster size = 10, than the intensity is 2.3266.

       

       While in slice viewer, if I set p = 0.01, cluster size = 10, than the intensity is 4.9267. Consequently, the significant clusters are far less than those in xjview. Why threshold should I adopt?  

       Thanks for helping. 

Submitted by ZHANG_RESTadmin on Tue, 11/12/2013 - 17:50

Xjview has the same default setting as SPM, that is, all your threshold p is single-tailed.

rest slice viewer, on the other hand, uses two-tailed p. 

So if you set p value in Xjview as 0.05 for positve activation only, you should set p value in REST as 0.1. 

Another issue is df setting, you will find different df correspond to different t threshold.