A bootstrap method for comparing correlated kappa coefficients. (English) Zbl 1431.62510

Vanbelle, S.; Albert, A.

J. Stat. Comput. Simulation 78, No. 11-12, 1009-1015 (2008).

Summary: Cohen’s kappa coefficient is traditionally used to quantify the degree of agreement between two raters on a nominal scale. Correlated kappas occur in many settings (e.g., repeated agreement by raters on the same individuals, concordance between diagnostic tests and a gold standard) and often need to be compared. While different techniques are now available to model correlated \(\kappa \) coefficients, they are generally not easy to implement in practice. The present paper describes a simple alternative method based on the bootstrap for comparing correlated kappa coefficients. The method is illustrated by examples and its type I error studied using simulations. The method is also compared with the generalized estimating equations of the second order and the weighted least-squares methods.

MSC:

62P10

Applications of statistics to biology and medical sciences; meta analysis

Keywords:

Cohen’s kappa; comparison; bootstrap

References:

[1]	Cohen J., Educational and Psychological Measurement 20 pp 37– (1960) · doi:10.1177/001316446002000104
[2]	Fleiss J. L., Statistical methods for rates and proportions, 2. ed. (1981) · Zbl 0544.62002
[3]	McKenzie D. P., Journal of psychiatric research 30 pp 483– (1996) · doi:10.1016/S0022-3956(96)00033-7
[4]	Williamson J. M., Biostatistics 1 pp 191– (2000) · Zbl 0959.62110 · doi:10.1093/biostatistics/1.2.191
[5]	Lipsitz S. R., Journal of the Royal Statistical Society A 164 pp 449– (2001) · Zbl 1002.62523 · doi:10.1111/1467-985X.00213
[6]	Barnhart H. X., Biometrics 58 pp 1012– (2002) · Zbl 1210.62142 · doi:10.1111/j.0006-341X.2002.01012.x
[7]	Efron B., An introduction to the bootstrap (1993) · Zbl 0835.62038
[8]	Kraemer H. C., Psychometrika 44 pp 461– (1979) · Zbl 0425.62088 · doi:10.1007/BF02296208
[9]	Cohen J., Psychological bulletin 70 pp 213– (1968) · doi:10.1037/h0026256