Authors

Ge Lin

Document Type

Working Paper

Publication Date

2002

College/Unit

Regional Research Institute

Document Number

Research Paper #2002-13

Department/Program/Center

Regional Research Institute

Abstract

In this paper, I propose a logit spatial association model for binary spatial events and develop a scan algorithm to search for spatial associations. I extend the traditional logit model with spatial autocorrelated component so that the model includes not only known risk factors, but also spatially autocorrelated regions as a control or explanatory factor. The case study of West Virginia Lung Cancer shows that the model effectively captures cool and hot spots in lung cancer mortality.

Share

COinS