Resumen de Bayesian inference for modelling the uncertainty by a mixture model for rating data

Maria Iannario, Claudia Tarantola

• In this paper we perform Bayesian quantitative analysis of the cup model, which is a two-component mixture model recently introduced for the analysis of ordinal data. It combines a standard cumulative model with a discrete Uniform distribution used to take into account the uncertainty in the rating process.

  Since the posterior distribution of the parameters of interest is not in a closed form, MCMC methods are used to simulate from it. The performance of the proposed methodology has been evaluated via real data o ering practical suggestions for using this approach in social-science, medicine or economic settings when an ordinal response is provided.