calculating weighted kappa statistics

[Maggie Cheang <mcheang@mac.com>]

Hi, I would like to calculate a kappa statistics for two tests that  
rate the same samples on a scale of 0,1,2 and 3.
However, I would like to penalize less if one test gives 2 and the  
other test gives 3, in short, the closer the "scale" of the test,  
less penalty.
i am wondering how I can achieve this in s-plus?
thank you very much

Maggie Cheang
mcheang@mac.com