fleiss.pair: Compare G independent pairwise kappa coefficients

View source: R/multiagree.R

fleiss.pairR Documentation

Compare G independent pairwise kappa coefficients

Description

This functions compare L independent kappa coefficients using the method of Fleiss (1981)

Usage

fleiss.pair(rater1, rater2, cluster_id, weight = "equal", multilevel = T,
  a.level = 0.05, cov, ITN = 1000, meth)

Arguments

rater1

a vector with the ratings of the first observer

rater2

a vector with the ratings of the second observer

cluster_id

a vector with the identification number of the clusters

weight

the weighting scheme to be used for kappa coefficients. 'unweighted' for Cohen's kappa, 'equal' for linear weights and 'squared' for quadratic weights

multilevel

a binary indicator equal to TRUE if the data are multilevel and FALSE otherwiwse.

a.level

the significance level

cov

the covariate determining the L groups to be compared

ITN

number of bootstrap iterations needed if the bootstrap procedure is chosen

meth

the method to be used to compute the standard error of the kappa coefficients: 'delta' for the delta method and 'boot' for the bootstrap method.

Details

This function compare L independent kappa coefficients using the method of Fleiss (1981). The data have to be entered in a vertical format.

Value

$kappa the value of the L kappa coefficients and their standard error

$chi the value of the chi-squared statistic with L-1 degrees of freedom

$p the p-value

Author(s)

Sophie Vanbelle sophie.vanbelle@maastrichtuniversity.nl

References

Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.). New York: John Wiley

Vanbelle S. Comparing dependent agreement coefficients obtained on multilevel data. submitted

Vanbelle S. (2014) A New Interpretation of the Weighted Kappa Coefficients. Psychometrika. Advance online publication. doi: 10.1007/s11336-014-9439-4

Examples

 
 
#dataset (multilevel) (Vanbelle, xxx)

data(FEES)
attach(FEES)

fleiss.pair(rater1=val_CO,rater2=val_COR,cluster_id=subject,
weight="unweighted",multilevel=TRUE,meth='delta',cov=group)

svanbelle/multiagree documentation built on Feb. 9, 2023, 2:37 p.m.