|
Abstract:
|
Automated geocoding, which attempts to match each subject's address to an address-ranged street segment georeferenced within a streetline database and then interpolates the position of the address along that segment, is an important data assimilation step in many geographic public health studies. Unfortunately, geocoding may result in many positional errors. Our study sought to model the probability distribution of positional errors associated with automated geocoding (both 100%-matched and 60%-matched) and E911 geocoding for data from a rural health study in Iowa. We found mixtures of bivariate t distributions with few components to be flexible enough to fit these data, yet parsimonious enough to be feasible for nascent applications of measurement-error methodology to spatial epidemiology. We conjecture that bivariate t mixtures will generally be useful for geocoding errors.
|
