4323.0: Tuesday, October 23, 2001: 4:30 PM-6:00 PM | ||||
Oral Session | ||||
| ||||
| Geographic information has long played a crucial role in helping to elucidate disease etiology and to clarify the impacts of on population health of climate, pollutants of air and water, and other environmental factors. However, biostatistical modeling of geographic and spatial data, and formal inferences from resulting models, have been difficult because i) shared influences, not all of which may be anticipated and measured, can induce substantial correlations between observations on geographical units, and ii) these correlations may be expected to vary in relation to general measures of the extent of shared influences, of which physical proximity is the most obvious. In this sense, geographic data share certain features of time series, but with a more ambiguous measure of distance between potentially correlated observations. Advances in computation and associated work in Bayesian analyses have produced an explosion in methodology and software for the presentation and analysis of geographical information, including geographic information collected longitudinally. Some of this work has been commercially packaged as "geographic information systems," while other work remains as yet within the academic community, undergoing continuing development and refinement. This session describes and illustrates the application of modern tools for exploration and analysis of geographic and geo-temporal information. Two papers concern methods for hypothesis-based exploration and modeling public health surveillance data, while the third specifically considers tools for using geographical data to study questions of equity in the development, implementation, and enforcement of environmental laws, regulations and policies. | ||||
| See individual abstracts for presenting author's disclosure statement. | ||||
| Learning Objectives: At the conclusion of the session, the participant will be able to: 1. Recognize the manner in which geographic proximity may introduce correlation into spatial data, and the manner in which unacknowledged correlation may distort standard statistical analyses. 2. Identify general methodological approaches to adjusting for spatial correlation in the exploration and formal analysis of spatial data. 3. List several tools for correlation-adjusted analysis of geographic data, and describe their application to public health surveillance data and questions of environmental equity. | ||||
| Bradley P. Carlin, PhD | ||||
| Bradley P. Carlin, PhD | ||||
| Introductory Remarks | ||||
| Exploratory Spatial and Temporal Analysis for Public Health Surveillance Data Owen Devine | ||||
| Linking GIS and statistical tools to assess environmental justice Lance A. Waller, PhD, Andrew B. Barclay | ||||
| Frailty Modeling for Spatially Correlated Survival Data, with Application to Infant Mortality in Minnesota Bradley P. Carlin, PhD | ||||
| Discussion | ||||
| Sponsor: | Statistics | |||
| Cosponsors: | Environment | |||
| CE Credits: | CME, Chiropractic, Environmental Health, Health Education (CHES), Nursing, Pharmacy, Social Work | |||