Measuring Agreement for Multinomial Data. Davies, M. & Fleiss, J L Biometrics, 38(4):1047–1051, December, 1982.
Measuring Agreement for Multinomial Data [link]Paper  doi  abstract   bibtex   
A kappa-like statistic is proposed as a measure of agreement for a set of multinomial random variables arrayed in a two-way laytout. This statistic is shown to arise either via the chance-correction of a certain proportion of agreement or via an analysis of variance for a two-way layout. The asymptotic standard error is derived under the assymption of complete independence among the random variables.
@article{davies_measuring_1982,
	title = {Measuring {Agreement} for {Multinomial} {Data}},
	volume = {38},
	copyright = {Copyright © 1982 International Biometric Society},
	issn = {0006-341X},
	url = {http://www.jstor.org/stable/2529886},
	doi = {10.2307/2529886},
	abstract = {A kappa-like statistic is proposed as a measure of agreement for a set of multinomial random variables arrayed in a two-way laytout. This statistic is shown to arise either via the chance-correction of a certain proportion of agreement or via an analysis of variance for a two-way layout. The asymptotic standard error is derived under the assymption of complete independence among the random variables.},
	number = {4},
	urldate = {2014-10-26},
	journal = {Biometrics},
	author = {Davies, Mark and Fleiss, J L},
	month = dec,
	year = {1982},
	keywords = {kappa, null},
	pages = {1047--1051},
}

Downloads: 0