ISEE-0108 Background and Objective: Although sources of error in geographic locations (e.g. geocoding error) used for describing and modeling spatial disease patterns are widely acknowledged, research on how such error impacts the results has been limited. In this paper we explore techniques for quantifying the perturbability of spatial weights to different specifications of geocoding error. Methods: New measures for assessing impacts of geocoding error on spatial weights called leverage and perturbability are used to construct the LIGE (Local Indicators of Geocoding Error) scatterplot. We evaluate the approach in simulation studies, and then demonstrate it using a case-control study of bladder cancer in south-eastern Michigan. Results: The LIGE scatterplot may be used to evaluate sensitivity of alternative spatial weight specifications to geocoding error both globally (when all locations are considered simultaneously) and locally (to identify those locations that would benefit most from increased geocoding accuracy). Conclusion: Three results are significant. First, the shape of the geocoding error functions we explored (e.g. circular, elliptical, greek cross) is not important; what matters is the mean geocoding error and its variance. Second, our methodology allows researchers to evaluate the sensitivity of spatial statistics to geocoding error for specific geographies. This has substantial practical implications since it makes possible routine sensitivity analysis of spatial statistics to geocoding error. Third, those locations with high perturbability (most sensitive to geocoding error) and high leverage (that contribute the most to the spatial weight specification being considered) will benefit the most from increased geocoding accuracy, and are rapidly identified using the new visualization tool we call the LIGE scatterplot.
Read full abstract