Yes, it is possible. The max p-value is just the p-value for a vertex. It is the probably of getting the data at that vertex by chance. The p-value for the cluster is the probably of getting a cluster of the given size by chance. The two p-values are unrelated. doug
On 05/07/2014 09:45 AM, Las Blawimo wrote:
Dear sir, madam, I am currently using Freesurfers Qdec and I have a question about the statistical output: is it possible to have a peak p-value that is larger (i.e. less significant) than the clusterwise p-value? My peak p-value (after using a - log10 transformation) = 0.0038, while my clusterwise p-value = 0.0029. Could you maybe give an explanation how the clusterwise p-value is computed? Thank you very much! Kind regards, Las Blawimo
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer