Title: Bayesian two-part model for semicontinuous data with latent variables
Authors: Xiaoqing Wang - The Chinese University of Hong Kong (Hong Kong) [presenting]
Xinyuan Song - Chinese University of Hong Kong (Hong Kong)
Abstract: A joint modeling approach is proposed to investigate the observed and latent risk factors of semicontinuous responses of interest. The proposed model consists of two major components. The first component is a structural equation model (SEM), which characterizes latent variables through multiple observed variables and simultaneously assesses interrelationships among the latent variables. The second component is a two-part model for investigating the effects of observable and latent variables on semicontinuous responses of interest. The two-part model comprises a model for a binary indicator variable and a model for another response variable that is conditioned on the binary indicator variable. A full Bayesian approach with Markov chain Monte Carlo algorithm is developed for statistical inference. A simulation study demonstrates the satisfactory performance of the developed methodology. The method is then applied to a study concerning the relationship among non-cognitive abilities, education, and annual income.