[R] Partek has Dunn-Sidak Multiple Test Correction. Is this the same/similar to any of R's p.adjust.methods?

Martin Maechler maechler at stat.math.ethz.ch
Fri Jul 15 15:27:56 CEST 2005


>>>>> "Earl" == Earl F Glynn <efg at stowers-institute.org>
>>>>>     on Thu, 14 Jul 2005 12:22:49 -0500 writes:

    Earl> The Partek package (www.partek.com) allows only two selections for Multiple
    Earl> Test Correction:  Bonferroni and Dunn-Sidak.  Can anyone suggest why Partek
    Earl> implemented Dunn-Sidak and not the other methods that R has?  Is there any
    Earl> particular advantage to the Dunn-Sidak method?
    Earl> R knows about these methods (in R 2.1.1):

    >> p.adjust.methods
    Earl> [1] "holm" "hochberg" "hommel" "bonferroni" "BH" "BY" "fdr"
    Earl> [8] "none"

    Earl> BH is Benjamini & Hochberg (1995) and is also called "fdr" (I wish R's
    Earl> documentation said this clearly).  BY is Benjamini & Yekutieli (2001).

The current R docu has

  >>   The '"BH"' and '"BY"' method of Benjamini, Hochberg, and Yekutieli
  >>   control the false discovery rate, the expected proportion of false
  >>   discoveries amongst the rejected hypotheses.  The false discovery
  >>   rate is a less stringent condition than the family wise error
  >>   rate, so these methods are more powerful than the others.

so both "BH" and "BY"   ``are FDR versions''. 
"fdr" was used - unfortunately - in some older versions of R,
so we kept it working as an *alias* for the time being.  
You should rather not know about it :-)
and use "BH" or "BY" (and maybe other methods in the future) instead.

Regards,

Martin




More information about the R-help mailing list