Spatial Autocorrelation in Regression Analysis: A Beginner’s Guide

Since about 1960, there has been a rapidly increasing technical literature on the concept of spatial autocorrelation and its implications for statistical estimation in geography. Because this literature has tended, of necessity, to be theoretical and difficult to follow, many potential users have either shied away from it or misunderstood its usefulness and its limitations. This is unfortunate. Many empirical geographers could beneficially use the techniques developed in this area, especially in regression analysis with zonal data. Further, this literature identifies a number of fundamental conceptual issues in geographic research of which every geographer should be aware. It is with these concerns in mind that I have written this paper. The purpose of the paper is to provide a relatively non-technical introduction to spatial atuocorrelation and its implications for regression analysis.