[R] what does the "S.D." returned by {Hmisc} rcorr.cens measure?

[Frank Harrell]

Dxy is a U-statistic.  In the U-statistic literature there is a combinatoric
approach to estimating variances, requiring one to examine all possible
pairs of observations.  The general formula I use is a bit messy.  You can
look at the Fortran code that comes with the Hmisc package to see the
algorithm.
Frank


vikkiyft wrote:
> 
> Thanks for your reply Prof Harrell!!
> 
> Could you kindly list some references of the fomula for calculating the SD
> of Somer's D in this kind of application? Because I couldnt find any..
> 


-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context: http://r.789695.n4.nabble.com/what-does-the-S-D-returned-by-Hmisc-rcorr-cens-measure-tp3329609p3331566.html
Sent from the R help mailing list archive at Nabble.com.