p-value > 1

Thank you for your reply.

The only contact address of the maintainer is
http://www-stat.stanford.edu/~tibs/SAM

However, one find no contact when one visit this page.

I am not a mathematician, therefore, it is strange for me to get probability bigger that 1, if it is like that should I trust the p-value of 0.004332?

Sorry for these questions.

Best wishes,
Amor


--- Peter Dalgaard <p.dalgaard at biostat.ku.dk> schrieb am Fr, 13.11.2009:

Von: Peter Dalgaard <p.dalgaard at biostat.ku.dk>
Betreff: Re: [R] p-value > 1
An: "amor Gandhi" <amorigandhi at yahoo.de>
CC: r-help at stat.math.ethz.ch
Datum: Freitag, 13. November 2009, 15:24

amor Gandhi wrote:

Dear all,

I am trying to use SAMr-library(samr), it gives me p-value = 1.001, any idea why?

Well, the package maintainer might be a better place to ask. (Bring an example if you do.)

However, as a generic matter, the result is obviously nonsensical, and could be due to a sloppy approximation of a distribution function, or maybe it is a (say) Bonferroni-adjusted p-value which is really an upper bound for the correct p-value.

Notice, though, that you should only very rarely care about the exact p-value. As soon as it is above 0.1, the message is just that you have not proven that there is an effect.

--? ? O__? ---- Peter Dalgaard? ? ? ? ? ????ster Farimagsgade 5, Entr.B
? c/ /'_ --- Dept. of Biostatistics? ???PO Box 2099, 1014 Cph. K
?(*) \(*) -- University of Copenhagen???Denmark? ? ? Ph:? (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)? ? ? ? ? ? ? FAX: (+45) 35327907