Same p-value threshold for FDR and bonferonni correction

[bong516612]

Dear all,

I am trying to perform paired t- test for my connectivity matrix.
To adress multiple correction problem, I use permutation based paired t test (permutation=1000).
After that, when I adjust FDR and bonferroni, I got strange message in command.

BST> Average corrected p-threshold: 0.000153846 (Bonferroni, Ntests=325)
BST> Average corrected p-threshold: 0.000153846 (FDR, Ntests=325)

why 2 corrected p-threshold is same? I think 0.05 / 325 = 0.000153846, but, isn't FDR usually showing higher p-threshold than Bonferroni? I don't know whats happening in here!

[SBeumer]

@bong516612

I am trying to perform paired t- test for my connectivity matrix.
To adress multiple correction problem, I use permutation based paired t test (permutation=1000).
After that, when I adjust FDR and bonferroni, I got strange message in command.

BST> Average corrected p-threshold: 0.000153846 (Bonferroni, Ntests=325)
BST> Average corrected p-threshold: 0.000153846 (FDR, Ntests=325)

why 2 corrected p-threshold is same? I think 0.05 / 325 = 0.000153846, but, isn't FDR usually showing higher p-threshold than Bonferroni? I don't know whats happening in here!

Both corrections can give approximately the same value under very specific circumstances as shown in this explanation on FDR (https://matthew-brett.github.io/teaching/fdr.html)