Re: [S] strange p-value for Fisher exact's test

Vadim A Kutsyy (kutsyy@umich.edu)
Tue, 24 Mar 1998 09:58:06 -0500


I used to work at software testing (for one of the stat software
company). And all programs have there bugs, with some strange data
sets, on some systems.
Any software is system dependent up to some point.

I would recommend to email this to mathsoft and they will fix it.

PS: There is a data set for SPSS 6.0 which would give 130% of value
count, another data set for SAS 6.? which would give mean larger than
max. value

Arnold.Dekkers@rivm.nl wrote:
>
> Very strange enough I get the same result als marcia:
>
> fisher.test(rbind(c(1, 1), c(71, 77)))
>
> Fisher's exact test
>
> data: rbind(c(1, 1), c(71, 77))
> p-value = 1.0001
> alternative hypothesis: two.sided
>
> This results comes from
> * beta-version S-PLUS 4.5 and
> * from S-PLUS 4.0 release 2
> * and from Version 3.3 Release 1 for MS Windows 3.1 : 1995
>
> under Windows NT on a Acer with Pentium processor.
>
> So, what's going on ? Are the S-PLUS results machine dependent ??
>
> Arnold
>
> Arnold L.M. Dekkers Ph.D.
>
> National Institute for Public Health and the Environment
>
> RIVM
> P.O. BOX 1
> NL 3720 BA BILTHOVEN
> THE NETHERLANDS
>
> tel. +31 - 30 - 274 23 66
> fax. +31 - 30 - 274 44 27
>
> email : Arnold.Dekkers@rivm.nl
>
> -----------------------------------------------------------------------
> This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
> send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
> message: unsubscribe s-news

-- 
Vadim Kutsyy
kutsyy@umich.edu
http://www.stat.lsa.umich.edu/~kutsyy
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu.  To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message:  unsubscribe s-news