[R] Cohen's Kappa

[Scot W McNary]

Hi,

Chance is not .5 in your data, it's a function of the expected values 
for presence and absence:

> (((7792*10855)/11974) + ((4182*1119)/11974))/11974
[1] 0.6225686

> (.6862368-.6225686)/(1-.6225686)
[1] 0.1686881

Scot


On Thu, 22 Mar 2007, Christian Schulz wrote:

> Hi,
>
> im little bit confused about Cohen's Kappa and i should  be look into the
> Kappa function code. Is the easy formula really wrong?
>
> kappa=agreement-chance/(1-chance)
>
> many thanks
> christian
>
> ###############################################################################
> true-negativ:7445
> false-positive:3410
> false-negativ:347
> true-positiv:772
>
> classification-aggrement:68,6%
> kappa=agreement-chance/(1-chance) = (0.686-0.5)/0.5=0.372
>
> .....with function from library(vcd)
> Kappa(matrix(c(7445,3410,347,772),nrow=2))
>               value         ASE
> Unweighted 0.1686882 0.011235188
> Weighted   0.1686882 0.007979293
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


--
   Scot W. McNary  email: smcnaryatcharmdotnet