Estimation of discrete/continuous choice model: application of Bayesian approach using Gibbs sampling

Abstract

The paper suggests a different model for estimating a discrete/continuous choice framework using Gibbs sampling. The basic idea of using Gibbs sampling for the multinomial probit model is extended to the case of a discrete/continuous framework. To check the applicability of this method, it was applied to estimate the parameters of a ‘vehicle ownership and use’ model in Seoul, Korea. Also estimated was the price elasticity of gasoline demand, calculated to be -0.7601.