Spatial Regression Analysis on the PC Spatial Statistics Using Sas

This monograph, part of the Resource Publications in Geography Series of the Association of American Geographers, presents standard statistical models modified to take into account the presence of spatial autocorrelation & SAS code for implementing these models. The first chapter defines the concept of spatial autocorrelation & discusses various interpretations of it. This chapter also presents two simulation games designed to familiarize players with spatial autocorrelation patterns. The second chapter reviews the basic statistical indices to measure levels of spatial autocorrelation in geographic distributions & presents SAS code for computing them using three sample data sets. The third & fourth chapters demonstrate how conventional regression models can be converted into one of three spatial autoregressive models. Detailed mathematical formulations of the models are given. SAS computer code for implementing each of the models is presented. The conclusion verifies the procedures described in the monograph by using them to estimate results reported in earlier empirical studies. Order from Association of American Geographers, 1710 16th St., N.W., Washington, D.C. 20009-3198; 202-234-1450.Griffith, Daniel A. is the author of 'Spatial Regression Analysis on the PC Spatial Statistics Using Sas' with ISBN 9780892912131 and ISBN 0892912138.