MULTIPARAMETRIC AND HIERARCHICAL SPATIAL AUTOREGRESSIVE MODELS: THE EVALUATION OF THE MISSPECIFICATION OF SPATIAL EFFECTS USING A MONTE CARLO SIMULATION

Abstract

The aim of this paper is to evaluate the spatial and hierarchical models for data generating processes with spatial heterogeneity and spatial dependence at the higher level. The simulation for the m-SAR and HSAR models was used to discuss the consequences of spatial misspecification. We noticed that the misspecification of spatial homogeneity or heterogeneity in both models affects i.a. the estimated parameter for spatial interactions at the individual level. Applying a m-SAR model for spatially heterogeneous processes causes the overestimation of the spatial interaction parameter.

Keywords

• Spatial model
• hierarchical model
• Monte Carlo, Bayesian estimation.