[R] exact p-value for Spearman, with ties

[Peter Dalgaard]

Paul H Artes <paul.h.artes at gmail.com> writes:

> Dear R,
> 
> I'm looking for exact p-values for Spearman's rank correlation in the
> presence of ties. This is available in StatXact and SPSS, but I
> haven't yet found it in R. Has anyone implemented this?

The algorithm used by those two (it's the same one -- SPSS licences
it) does not seem to be openly available. 

It should be fairly easy to simulate the distribution of rho.

Another matter is that we really ought to at least get the normal
approximation right in the presence of ties.


-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907