On 02/20/2014 04:33 AM, charujing123 wrote:
Hi FS experts and users, May I ask some questions about clusterwise correction for multiple comparisons? 1,In mri_glmfit-sim command, there is a --sim(--sim nulltype nsim threshold csdbase) option. Choose the voxel/vertex-wise threshhold used to define clusters. This will be -log10(p), where p is the p-value. Eg, to use all voxels/vertices with p < .01, then set the threshhold = 2.I want to know what is meaning of "using all voxels with p < 0.01". Does it mean that considering all voxels whose p value below 0.01 in every iterations. Then it generates some clusters in every iteration. Is it right?
At each iteration, it creates a random map, smooths it, then thresholds it to create clusters. It keeps track of how many clusters were created and the maximum cluster size at each iteration. It uses this info to compute a p-value for the clusters it finds in your real data.
2,Also there is a --sim-sign option in this command. I want to know meaning of the sign: abs,pos,neg.Does it represent one- and two-tailed tests?
Yes, abs is a two tailed, pos and neg are one-tailed either positive or negative tail doug
Thanks. All the best. Rujing Zha 2014-02-20
charujing123
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer