Evaluation Of Inter-Observer Reliability Of Animal Welfare Indicators: Which Is The Best Index To Use?

This study focuses on the problem of assessing inter-observer reliability (IOR) in the case of dichotomous categorical animal-based welfare indicators and the presence of two observers. Based on observations obtained from Animal Welfare Indicators (AWIN) project surveys conducted on nine dairy goat farms, and using udder asymmetry as an indicator, we compared the performance of the most popular agreement indexes available in the literature: Scott's pi, Cohen's k, k(PABAK), Holsti's H, Krippendorff's alpha, Hubert's Gamma, Janson and Vegelius' J, Bangdiwala's B, Andres and Marzo's Delta, and Gwet's gamma(AC(1)). Confidence intervals were calculated using closed formulas of variance estimates for pi, k, k(PABAK), H, alpha, Gamma, J, Delta, and gamma(AC(1)), while the bootstrap and exact bootstrap methods were used for all the indexes. All the indexes and closed formulas of variance estimates were calculated using Microsoft Excel. The bootstrap method was performed with R software, while the exact bootstrap method was performed with SAS software. k, pi, and alpha exhibited a paradoxical behavior, showing unacceptably low values even in the presence of very high concordance rates. B and gamma(AC(1)) showed values very close to the concordance rate, independently of its value. Both bootstrap and exact bootstrap methods turned out to be simpler compared to the implementation of closed variance formulas and provided effective confidence intervals for all the considered indexes. The best approach for measuring IOR in these cases is the use of B or gamma(AC(1)), with bootstrap or exact bootstrap methods for confidence interval calculation.