Flexible Bayesian Methodology for Multivariate Spatial Small Area Estimation

• Activity Number: 18
• Type: Topic Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Aaron T. Porter*+ and Scott Holan and Christopher K. Wikle
• Companies: University of Missouri and University of Missouri and University of Missouri
• Keywords: American Community Survey ; Areal Data ; Bayesian Hierarchical Model ; Multivariate

Abstract

The importance of explicitly accounting for spatial correlation in univariate small area estimation has recently been acknowledged. However, methodology for simultaneously and explicitly modeling within-area correlation and between-area spatial correlation in multivariate small area estimation is relatively under-developed. In this research, we develop hierarchical Bayesian methodology to flexibly handle explicit multivariate spatial dependence in the Fay-Herriot framework of models. The effectiveness of our approach is illustrated through simulation as well as data from the American Community Survey.