B0533
Title: Normalizing the weighted kappa in rater agreement problems
Authors: Fabio Rapallo - University of Genova (Italy) [presenting]
Abstract: In rater agreement analysis the computation of the maximum agreement given the margins is a crucial task in order to obtain correctly normalized indices. The notion of the Markov move from algebraic statistics is used to analyze the weighted kappa indices. In particular, the problem of the maximum kappa and its dependence on the choice of the weighting schemes are discussed. The Markov moves are also used in a simulated annealing algorithm to actually find the configuration of maximum agreement. Finally, an alternative approach to defining normalized kappa indices is discussed. This second approach is based on the theory of copulas and the iterative proportional fitting algorithm.
