A Deaggregating Model for the Semi-areal Data with a Bivariate CAR Prior

• Activity Number: 291
• Type: Topic Contributed
• Date/Time: Tuesday, August 3, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Ye Liang*+ and Dongchu Sun and Zhuoqiong He and Mario Schootman
• Companies: University of Missouri and University of Missouri and University of Missouri and Washington University in St. Louis
• Keywords: quality of life ; deaggregating model ; bivariate spatial model ; conditional autoregressive model ; Markov Chain Monte Carlo

Abstract

The quality of life (QOL) of cancer survivors is such a topic that is concerned by epidemiologists, medical care researchers and public health policy makers. Researchers hope to identify a geographical trend of the QOL. The quantitative evaluation is measured through survey questions, thus individual information available. The traditional data model is usually designed for the aggregated data. A deaggregating data model is proposed to retain the individual information. A conditional bivariate CAR structure is derived from the decomposition of Wishart random matrix, which is specified for the spatial effect. The setting has a clear interpretation, and is computationally accessible. The model can be applied by public health researchers without programming difficulties.