Regression analysis of spatially autocorrelated data : a study of county-level turnout in Texas

Date

2018-05

Authors

Gibson, Nadine Suzanne

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Interest in spatially weighted regression analysis has increased due to corresponding increases in access to publicly available spatial data. Spatial autocorrelation occurs when the ordering of observations across space produces a relationship between pairs of individual observations. Instances of spatial autocorrelation necessitate the use of alternative approaches to parameter estimation other than ordinary least squares. With a focus on autocorrelation resulting from spatial dependence in the dependent variable or the error term, this report summarizes basic methodology for detecting spatial autocorrelation and spatial autoregressive model selection. The approaches outlined in this report are then applied to an analysis of county-level turnout in Texas.

Department

Description

LCSH Subject Headings

Citation