Spatial Regression with Covariate Measurement Error: a Semiparametric Approach

• Activity Number: 231
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Md Hamidul Huque*+ and Howard D. Bondell and Raymond J. Carroll and Louise Ryan
• Companies: University of Technology and North Carolina State University and Texas A&M and University of Technology
• Keywords: Measurement error ; spatial correlation ; penalized likelihood ; splines ; thin plate spline regression basis

Abstract

Spatial data have become increasingly common in epidemiology and public health research due to the rapid advances in GIS (Geographic Information Systems) technology. In health research, for example, it is common for epidemiologists to incorporate geographically indexed data into their studies. In practice, however, the spatially-defined covariates are often measured with error. The classical measurement error theory is inapplicable in the context of spatial modeling because of the spatial correlation among the observations. The naïve estimator of regression coefficients are attenuated if measurement error is ignored. We proposed a semi parametric regression approach to obtain the bias corrected estimates of the regression parameter and derived the large sample properties of the estimates. We evaluate the performance of the proposed method through simulation studies and illustrate using real examples.