Dear FSL experts,
I have an analysis that sadly did not meet significance with TFCE.
Previous postings on the list said one can go as low as 1.5 for
cluster thresholding and do FDR as two acceptable approaches, however
I am not sure about using them together. When I use a cluster of 1.5
and do FDR correction I get nice results. But is this a valid approach?
Thanks,
Naama
|