The Impact of Weight Matrices on Parameter Estimation and Inference: A Case Study of Binary Response Using Land Use Data

This paper develops two new models and evaluates the impact of using different weight matrices on parameter estimates and inference in three distinct spatial specifications for discrete response. These specifications rely on a conventional, sparse, inverse-distance weight matrix for a spatial auto-regressive probit (SARP), a spatial autoregressive approach where the weight matrix includes an endogenous distance-decay parameter (SARPα), and a matrix exponential spatial specification for probit (MESSP). These are applied in a binary choice setting using both simulated data and parcel-level land use data. Parameters of all models are estimated using Bayesian methods. In simulated tests, adding a distance-decay parameter term to the spatial weight matrix did not alter the quality of estimation and inference, but the added sampling loop required to estimate the distance-decay parameter substantially increased computing times. By contrast, the MESSP model's obvious advantage is its fast computing time, thanks to elimination of a log-determinant calculation for the weight matrix. In the model tests using actual land use data, the MESSP approach emerged as the clear winner, in terms of fit and computing times. Results from all three models offer consistent interpretation of parameter estimates, with locations farther away from the regional central business district (CBD) and closer to roadways being more prone to (mostly residential) development (as expected). Again, the MESSP model offered the greatest computing-time savings benefits, but all three specifications yielded similar marginal effects estimates, showing how a focus on the spatial interactions and net (direct plus indirect) effects across observational units is more important than a focus on slope-parameter estimates, when properly analyzing spatial data.

[1]  Agostino Nobile,et al.  Comment: Bayesian multinomial probit models with a normalization constraint , 2000 .

[2]  J. LeSage.  Introduction to spatial econometrics , 2009 .

[3]  Samuel Labi,et al.  Contracting in highway maintenance and rehabilitation: Are spatial effects important? , 2010 .

[4]  Antonio Páez,et al.  Progress in Spatial Analysis , 2010 .

[5]  Luc Anselin,et al.  Under the hood , 2002 .

[6]  Kara M. Kockelman,et al.  Specification and estimation of a spatially and temporally autocorrelated seemingly unrelated regression model: application to crash rates in China , 2007 .

[7]  Kara M. Kockelman,et al.  The Dynamic Spatial Multinomial Probit Model: Analysis of Land Use Change Using Parcel-Level Data , 2011 .

[8]  Yiyi Wang.  ANTICIPATING LAND USE CHANGE USING GEOGRAPHICALLY WEIGHTED REGRESSION MODELS FOR DISCRETE RESPONSE , 2011 .

[9]  Molly Espey,et al.  Modeling Urban Sprawl and Land Use Change in a Coastal Area-- A Neural Network Approach , 2005 .

[10]  Yiyi Wang,et al.  Anticipation of Land use Change through use of Geographically Weighted Regression Models for Discrete Response , 2011 .

[11]  Kara M. Kockelman,et al.  Accessibility Indices: Connection to Residential Land Prices and Location Choices , 2002 .

[12]  L. Anselin.  Spatial Econometrics: Methods and Models , 1988 .

[13]  Frank Goetzke,et al.  Walkability as a Summary Measure in a Spatially Autoregressive Mode Choice Model: An Instrumental Variable Approach , 2010 .

[14]  Eric J. Miller,et al.  Effects of Transportation Infrastructure and Location on Residential Real Estate Values: Application of Spatial Autoregressive Techniques , 2000 .

[15]  Roger Th. A. J. Leenders,et al.  Modeling social influence through network autocorrelation: constructing the weight matrix , 2002, Soc. Networks.

[16]  Kara M. Kockelman,et al.  Neighborhood impacts on land use change: a multinomial logit model of spatial relationships , 2008 .

[17]  David B. Dunson,et al.  Bayesian Data Analysis , 2010 .

[18]  Nicholas J. Higham,et al.  The Scaling and Squaring Method for the Matrix Exponential Revisited , 2005, SIAM J. Matrix Anal. Appl..

[19]  Phillip Kostov,et al.  Model Boosting for Spatial Weighting Matrix Selection in Spatial Lag Models , 2010 .

[20]  Mark S. Mizruchi,et al.  The effect of density on the level of bias in the network autocorrelation model , 2008, Soc. Networks.

[21]  David M Levinson,et al.  Models of Transportation and Land Use Change: A Guide to the Territory , 2008 .

[22]  Alan Borning,et al.  Microsimulation of Urban Development and Location Choices: Design and Implementation of UrbanSim , 2003 .

[23]  John Geweke,et al.  Evaluating the accuracy of sampling-based approaches to the calculation of posterior moments , 1991 .

[24]  J. LeSage,et al.  The Biggest Myth in Spatial Econometrics , 2010 .

[25]  D. S. Bunch,et al.  Estimability in the Multinomial Probit Model , 1989 .

[26]  J. Ord,et al.  Spatial Processes. Models and Applications , 1982 .

[27]  Bayesian Estimation of A Distance Functional Weight Matrix Model , 2005 .

[28]  V. Chernozhukov,et al.  Bayesian Econometrics , 2007 .

[29]  Robert J. Franzese,et al.  Interdependence in Comparative Politics , 2008 .

[30]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[31]  Dominique Peeters,et al.  Estimation of the Coefficients of a Linear Regression in the Presence of Spatial Autocorrelation. An Application to a Belgian Labour-Demand Function , 1975 .

[32]  Steven Farber,et al.  Topology and Dependency Tests in Spatial and Network Autoregressive Models , 2009 .

[33]  James P. LeSage,et al.  A matrix exponential spatial specification , 2007 .

[34]  Debbie A. Niemeier,et al.  Accessibility: an evaluation using consumer welfare , 1997 .

[35]  Alan E. Gelfand,et al.  Spatial Modeling of House Prices Using Normalized Distance-Weighted Sums of Stationary Processes , 2004 .

[36]  A. Rukhin.  Bayes and Empirical Bayes Methods for Data Analysis , 1997 .

[37]  Duane F. Marble,et al.  Spatial Analysis: A Reader in Statistical Geography , 1968 .