Benjamini, Krieger and Yekutieli multiple comparisons correction

css

New Member
#1
Hi,
I am reporting the number of statistically significant differences found on a series of 110 t tests with and without applying the Benjamini, Krieger and Yekutieli multiple comparisons correction method. I have also performed a series of 110 linear regressions on the same DVs and I would like to also report these data with and without applying the same Benjamini, Krieger and Yekutieli correction. However, in the case of t tests, I could easily apply this method because it is the recommended option when performing multiple t tests in Prism software. However, this program does not provide this option when performing multiple regression. Does anyone know how could I do such correction?
 

j58

Active Member
#2
The Benjamini, Krieger and Yekutieli (BKY) procedure is easy to run "by hand." If you look up the original 2006 Biometrika paper, the procedure is pretty easy to understand. The R function p.adjust, which is included in a basic R installation, will do the Benjamini and Yekutieli (2001) adjustment, but I don't know of any R package that implements BKY. But you shouldn't shy away from looking up the original paper and just cranking it out by hand.
 
Last edited:

css

New Member
#3
Hi j58,

Thank you for your reply. You were right, I went to the original article and I was able to understand how to do the calculations myself.

Nevertheless, I would like to share that a researcher named N. Pike (in his article "Using false discovery rates for multiple comparisons in ecology and evolution") has kindly provided an excel routine that calculates this and other similar corrections (the results obtained with this script matched those I had obtained when doing my own calculations).