As part of my PhD project, I conducted a brain imaging study and ran multiple correlations. In order to control the false discovery rate, I used the Benjamini-Hochberg procedure and used a FDR of 5% (or Q = 0.05). I have an Excel spreadsheet with all the correlations I ran. I ordered all of my p values from smallest to largest, ranked them and used the formula (i/m)*Q to get my critical B-H value (where i is the rank of the individual p value, m is the total number of correlations I ran, and Q is 0.05). I found the largest p value that was smaller than its critical B-H value, and considered all of the correlations above it to be significant. For example, out of 72 correlations, 16 turned out to be significant with the Benjamini-Hochberg procedure (relative to 31 without any correction).

I would like to know how to report those findings in the results section of a manuscript I plan on submitting to a journal. Is it sufficient to describe my approach in my Methods section and report only the correlations that survived the correction with their actual p value? If not, what would be the best way to report my results?

Many thanks,

Valérie