Estimation of Regression Coefficients in the Presence of Spatially Autocorrelated Error Terms

TLDR

Spatial autocorrelation arises when individuals are related through geographic proximity. The study proposes a maximum‑likelihood method to jointly estimate the spatial correlation parameters and regression coefficients. The method includes a test for spatial autocorrelation and is demonstrated via a hedonic regression example. © 1988 MIT Press.

Abstract

Spatial autocorrelation occurs when population members are related through their geographic location. This paper presents a maximum likelihood procedure for simultaneously estimating the parameters of the correlation function and the regression coefficients. A test for the presence of spatial autocorrelation is also provided. Estimation of an hedonic regression illustrates the technique. Copyright 1988 by MIT Press.