Spatial autocorrelation

Abstract

The purpose of this paper is to present a test for spatial correlation of disturbance terms in regression models. The test is computationally simple, and does not require the model to be linear, the disturbance terms to be normally distributed, or panel data to be available. For purposes of illustration, the test is applied to a model of per capita county police expenditures. The results obtained are compatible with results obtained via the Moran I test.