手法を比較

	複数の評価者間の一致度を測るフリース・カッパ	コーエンのカッパ係数	統計的信頼性解析
分野	統計学	統計学	信頼性
系統	Hypothesis test	Hypothesis test	Regression model
提唱年	1971	1960	1998
提唱者	Joseph L. Fleiss	Jacob Cohen	William Meeker & Luis Escobar
種類	Non-parametric agreement measure	Inter-rater reliability coefficient	Parametric lifetime modeling
原典	Fleiss, J.L. (1971). Measuring Nominal Scale Agreement Among Many Raters. Psychological Bulletin, 76(5), 378–382.	Cohen, J. (1960). A Coefficient of Agreement for Nominal Scales. Educational and Psychological Measurement, 20(1), 37–46.	Meeker, W. Q., & Escobar, L. A. (1998). Statistical Methods for Reliability Data. Wiley. ISBN: 978-0-471-14328-4
別名	multi-rater kappa, Fleiss kappa, Fleiss' Kappa (Çoklu Değerlendirici Uyumu)	kappa coefficient, kappa statistic, Cohen's Kappa (Değerlendiriciler Arası Uyum)	Life Data Analysis, Survival Analysis (Engineering), Time-to-Failure Analysis, Güvenilirlik Analizi
関連	2	3	3
概要	Fleiss' Kappa is a non-parametric statistic for measuring the degree of agreement among three or more raters who classify items into mutually exclusive nominal categories. Introduced by Joseph L. Fleiss in 1971 as a generalization of Cohen's Kappa beyond two raters, it corrects observed agreement for the level of agreement expected by chance alone, making it the standard reliability index in medical diagnosis studies, content analysis, and multi-coder research.	Cohen's kappa (κ) is a statistical measure of inter-rater reliability for categorical classifications, introduced by Jacob Cohen in 1960. Unlike simple percent agreement, kappa corrects for the level of agreement that would be expected purely by chance, making it the standard metric when two raters independently assign observations to the same set of mutually exclusive categories.	Statistical reliability analysis models the time-to-failure of components, systems, or products using parametric lifetime distributions fitted to observed or censored failure data. Formalized comprehensively by William Q. Meeker and Luis A. Escobar in their 1998 Wiley monograph, the framework integrates maximum likelihood estimation, censoring mechanisms, and distributional diagnostics to produce probability-of-failure curves, hazard rates, and quantile estimates that support design, warranty, and maintenance decisions.