Bayesian analysis of a dynamic multivariate spatial ordered probit model

Spatial econometrics has few studies on multivariate ordinal responses. This study proposes a dynamic multivariate spatial ordered probit (DMSOP) model, which is the first attempt to capture temporal and spatial dependencies simultaneously for multivariate ordinal responses. The parameters are calculated using Bayesian inference based on Markov chain Monte Carlo sampling. The DMSOP model performs effectively with the simulated data. Furthermore, the DMSOP model is applied to two response variables, namely, the life satisfaction and self-rated health of adults in 25 provinces in China. The empirical results show that the model can effectively measure the spatial and temporal dependencies for multivariate ordinal responses.

Automated summary

This study proposes a dynamic multivariate spatial ordered probit (DMSOP) model, which is the first attempt to capture temporal and spatial dependencies simultaneously for multivariate ordinal responses. The parameters are calculated using Bayesian inference based on Markov chain Monte Carlo sampling. The empirical results show that the model can effectively measure the spatial and temporal dependencies for multivariate ordinal responses.