Suggestion for updating `p.adjust` with new method (BKY 2006)

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Suggestion for updating `p.adjust` with new method (BKY 2006)

Xiaoqing Claire Rong-Mullins
Dear R contributors,

I suggest adding a new method to `p.adjust` ("Adjust P-values for Multiple
Comparisons",
https://stat.ethz.ch/R-manual/R-devel/library/stats/html/p.adjust.html).

This new method is published in Benjamini, Krieger, Yekutieli 2016 Adaptive
linear step-up procedures that control the false discovery rate
(Biometrika). https://doi.org/10.1093/biomet/93.3.491

This paper described multiple methods for adjusting p-values, where the "TST"
method (Definition 6) performed the best when test statistics are
positively correlated, per my interpretation. This method can be labeled as
"BKY", for the three authors Benjamini, Krieger, Yekutieli.

I apologize if this is a duplication.


Best,
Claire

Xiaoqing Claire Rong-Mullins
Bioinformatic Specialist
Division of Biostatistics
College of Public Health
The Ohio State University

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Suggestion for updating `p.adjust` with new method (BKY 2006)

Martin Maechler
>>>>> Xiaoqing Claire Rong-Mullins
>>>>>     on Mon, 23 Jul 2018 12:03:55 -0400 writes:

    > Dear R contributors,
    > I suggest adding a new method to `p.adjust` ("Adjust P-values for Multiple
    > Comparisons",
    > https://stat.ethz.ch/R-manual/R-devel/library/stats/html/p.adjust.html).

    > This new method is published in Benjamini, Krieger, Yekutieli 2016 Adaptive
    > linear step-up procedures that control the false discovery rate
    > (Biometrika). https://doi.org/10.1093/biomet/93.3.491

    > This paper described multiple methods for adjusting p-values, where the "TST"
    > method (Definition 6) performed the best when test statistics are
    > positively correlated, per my interpretation. This method can be labeled as
    > "BKY", for the three authors Benjamini, Krieger, Yekutieli.

    > I apologize if this is a duplication.

Not at all a duplication.  AFAICS this is the first time this is
proposed here.  As a matter of fact this seems such a good (and well
confined) suggestion, that I've proposed it as semester project
here, and so with high probability we will get a "BKY" method
for p.adjust() within a few months.

Thank you, indeed, for the proposal!

Best,
Martin

--
Martin <[hidden email]>   http://stat.ethz.ch/~maechler
Seminar für Statistik, ETH Zürich   HG G 16         Rämistrasse 101
CH-8092 Zurich, SWITZERLAND
phone: +41-44-632-3408      fax: ...-1228     <><


    > Best,
    > Claire

    > Xiaoqing Claire Rong-Mullins
    > Bioinformatic Specialist
    > Division of Biostatistics
    > College of Public Health
    > The Ohio State University

    > [[alternative HTML version deleted]]

    > ______________________________________________
    > [hidden email] mailing list
    > https://stat.ethz.ch/mailman/listinfo/r-devel

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel