Spatial Uncertainty Estimation and Public Health Data

• Activity Number: 632
• Type: Invited
• Date/Time: Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Lance A. Waller, Ph.D.*+ and Howard Chang and John Pearce and Matthew Strickland and Stefanie Sarnat and Paige Tolbert
• Companies: Emory University and Emory University and Emory University and Emory University and Emory University and Emory University
• Keywords: Spatially varying coefficients ; Geographically weighted regression ; Spatial Models ; Air pollution epidemiology

Abstract

Epidemiologic studies of the health effects of air pollution involve multiple sources of spatial uncertainty including uncertainties in exposures and their surrogates, uncertainties in locations of individuals (home or work), and uncertainties in estimated health effects. Due to strong seasonal trends and sharp temporal variations in exposures and outcomes, air pollution epidemiology has given careful attention to temporal variations but less so to the impact of spatial uncertainties on inference for the associations of interest. In this presentation, we build on ongoing air pollution epidemiology projects in Atlanta, Georgia to provide an overview of spatial uncertainties in exposures, outcomes, and the associations between the two with particular emphasis on a comparison between geographically weighted regression and model-based spatially varying coefficient models.