I'm now writing a paper, and in it I'll be using a certain FDR (False Discovery Rate) statistical procedure. It's a clever and not-too-conservative way to adjust for multiple comparisons, and to keep P-values in check. You should absolutely use it in your work if you have not being doing it yet:
But what I don't quite understand here is whether I need to reference the original papers in which the FDR method was described for the first time, or not.
The method is not too old: both papers that justify it were published in 1995 (see 2 first references in the Wikipedia article). At the same time by now this method is kind of known, and used, and based on the Google Scholar statistics the firsts paper of these two has an impressive number of 15582 citations. That's a lot! Does it mean that I can afford not referencing it?
Also would the fact that the Wikipedia article is that nicely written, and comes as the first result in Google Search, affect my decision?
Generally, what are the criteria? When do you stop referencing a methodological paper like this one?
