Bayesian Approach to Zero-Inflated Ordered Probit Models

• Activity Number: 33
• Type: Contributed
• Date/Time: Sunday, August 1, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Getachew Dagne*+
• Companies: University of South Florida
• Keywords: Ordinal model ; zero-inflation ; hierarchical model ; Bayesian inference

Abstract

This paper presents a Bayesian analysis of ordinal data with excess of zeros. The zero inflation and ordered choices are modeled using correlated latent variables structure and implemented in ordered probit paradigm. The estimation of parameters is carried out using Markov Chain Monte Carlo techniques. Using household tobacco survey data with substantial proportion of zeros, we analyze the socio-economic determinants of individual problem of smoking tobacco. In our illustration, we find strong evidence that accounting for excess zeros provides good fit to the data. The example shows that the use of a model that ignores zero-inflation masks differential effects covariates have on the two regimes, non-users versus users at all levels of consumption.