fleiss_k: Fleiss' kappa

In lbraglia/lbagree: Agreement miscellaneous utilities

Fleiss' kappa

Description

Fleiss' kappa (Cohen's equivalent for multiple raters)

Usage

fleiss_k(
  x,
  nlevels = NULL,
  raters_par = list(test = "MC", B = 1000, alpha = 0.05),
  irr_par = list(exact = FALSE, detail = FALSE),
  gwet_par = list(weights = "unweighted", categ = NULL, conflev = 0.95, N = Inf)
)

Arguments

x	a data.frame with one column per scale item/subject and one row per rater. Judgement can be a factors or integers
nlevels	number of levels of possible rating/judgement
raters_par	arguments passed to raters::concordance
irr_par	arguments passed to irr::kappam.fleiss
gwet_par	arguments passed to irrCAC::gwet.ac1.dist

Examples

## reproducing the example from
## https://en.wikipedia.org/wiki/Fleiss%27_kappa
item1  <- rep(5, 14)
item2  <- c(rep(1,0), rep(2,2), rep(3,6), rep(4,4), rep(5,2))
item3  <- c(rep(1,0), rep(2,0), rep(3,3), rep(4,5), rep(5,6))
item4  <- c(rep(1,0), rep(2,3), rep(3,9), rep(4,2), rep(5,0))
item5  <- c(rep(1,2), rep(2,2), rep(3,8), rep(4,1), rep(5,1))
item6  <- c(rep(1,7), rep(2,7), rep(3,0), rep(4,0), rep(5,0))
item7  <- c(rep(1,3), rep(2,2), rep(3,6), rep(4,3), rep(5,0))
item8  <- c(rep(1,2), rep(2,5), rep(3,3), rep(4,2), rep(5,2))
item9  <- c(rep(1,6), rep(2,5), rep(3,2), rep(4,1), rep(5,0))
item10 <- c(rep(1,0), rep(2,2), rep(3,2), rep(4,3), rep(5,7))

df <-  data.frame(item1, item2, item3, item4, item5,
                  item6, item7, item8, item9, item10)
df
fleiss_k(x = df, nlevels = 5)

lbraglia/lbagree documentation built on July 3, 2023, 3:32 p.m.