A model for analyzing rank data obtained from multiple evaluators, possibly using different ranking criteria, is proposed. The model is specified hierarchically within the Bayesian paradigm and includes parameters that represent the probabilities that two items are assigned equal rankings. Also included are parameters that account for the relative precision of rankings obtained from distinct evaluation schemes. The model is illustrated through a meta-analysis of rank data collected for the purpose of comparing the cognitive abilities of various primate genera.
The full text is available in PostScript format.