Title: Bayesian Estimation of Kendall's tau Using a Latent Normal Approach

Abstract: The rank-based association between two variables can be modeled by introducing a latent normal level to ordinal data. We demonstrate how this approach yields Bayesian inference for Kendall's rank correlation coefficient, improving on a recent Bayesian solution from asymptotic properties of the test statistic.