Dear TFCE users,
I have two questions about the interpretation of TFCE results using the SPM TFCE toolbox by Christian Gaser.
My first question is about the interpretation of the absolute TFCE values. I have performed TFCE analysis twice for two separate studies and once I got TFCE values in the range of 10.000 - 20.000, while the other time they were in the range of 100-1000 and even one peak voxel of a cluster (significant with FDR) had a TFCE score of 30.
Did something perhaps go wrong or is this perfectly possible? If it is possible, does anyone know in what way I should take into account the TFCE scores for interpreting the results?
A problem which was mentioned a time ago in this mailing list, was that the TFCE analysis of someone's study showed nearly all voxels of the brain as significant. I have this same issue now. However, in my case, there is clearly one cluster with a much higher TFCE score, but the cluster is so large, it takes nearly all voxels of the brain. The response to this problem earlier was to use a stricter threshold, but in my case either the entire cluster is significant or it isn't when using different thresholds. From what I understand, you're not supposed to look for the right threshold in order that you have nicely separated clusters, or would it be acceptable in this case?
Thanks in advance,
Steven
|