Hi,
I don't quite understand the p-value-correction for cluster-levels.
We are using SPM 8 r4467 (no additional toolboxes for the statistical output) with defaults settings:
defaults.stats.topoFDR = 1;
defaults.stats.rft.nonstat = 0;
When I use "p value adjustment to control" =FWE with threshold 0.05 the cluster p-values behave in a manner that I didn't expected:
p_FWE > p_uncorr
Does anyone can give me a hint where I can find some background information on the cluster p-values?
Thank you very much in advance!
Best regards,
Katharina
|