Title: Hierarchical models for rater agreement and the evergreen kappa statistic
Authors: Mauro Gasparini - Politecnico di Torino (Italy) [presenting]
Abstract: The kappa statistic as a measure of rater agreement will turn 60 soon and yet some of its properties are still attracting a lot of interest among statisticians. In particular, the appropriate modeling of rater agreement at the population level and the use of kappa for inferential purposes have been studied by several authors with different approaches. Some of the asymptotic distributions of kappa, an approach based on hierarchical models and their use in clinical and psychological research will be presented.