We would like to compare the inter-rater reliability of 7 raters over a gold standard. Our study's outcome variable has 3 categories. So, I have tried calculating the sample size using the package "kappaSize" and CI3cats function.
a<-CI3Cats(kappa0=0.8, kappaL=0.7, kappaU=0.9, props=c(0.5, 0.3, 0.2), raters=6, alpha=0.05);
summary(a)
This accepts only a maximum of 6 raters and produces an error when I try to enter raters=7.
Error in CI3Cats(kappa0 = 0.8, kappaL = 0.7, kappaU = 0.9, props = c(0.5, : Sorry, this function is designed for between 2 to 6 raters.
Is there a solution for calculating the sample size for comparing 7 raters over a gold standard?
