Accounting for Locational, Temporal, and Physical Similarity of Residential Sales in Mass Appraisal Modeling: Introducing the Development and Application of Geographically, Temporally, and Characteristically Weighted Regression (GTCWR)

Paul Bidanset, Michael McCord, John A Lombard, Peadar Davis, William McCluskey

Research output: Contribution to journalArticle

Abstract

Geographically weighted regression (GWR) has been recognized in the assessment community as a viable automated valuation model (AVM) to help overcome, at least in part, modeling hurdles associated with location, such as spatial heterogeneity and spatial autocorrelation of error terms. Although previous researchers have adjusted the GWR weights matrix to also weight by time of sale or by structural similarity of properties in AVMs, the research described in this paper is the first that has done so by all three dimensions (i.e., location, structural similarity, and time of sale) simultaneously. Using 24 years of single-family residential sales in Fairfax, Virginia, we created a new locally weighted regression (LWR) AVM called geographically, temporally, and characteristically weighted regression (GTCWR) and compared it with GWR-based models with fewer weighting dimensions.GTCWR was the only model to achieve IAAO-accepted levels of the coefficient of dispersion (COD), price-related differential (PRD), price-related bias (PRB), and median assessment-to-sale price ratio in both the training and testing samples, although it did not fully correct the existence of heteroscedasticity. With lower PRD and PRB levels, the application of temporal weighting to this data set did appear to help reduce indicators of vertical inequity. Along with an equitable, uniform, and defensible methodology that mirrors the sales comparison, GTCWR presents a new AVM that demonstrates an ability to value over 24 years of sales at IAAO standard levels, without the creation and implementation of time-based variables, the trimming of outliers, and time-intensive model specification and calibration
LanguageEnglish
Pages5-13
JournalJournal of Property Tax Assessment and
Volume14
Issue number2
Early online date31 Jan 2018
Publication statusE-pub ahead of print - 31 Jan 2018

Fingerprint

valuation
modeling
outlier
autocorrelation
appraisal
price
calibration
matrix
methodology

Keywords

  • Property Tax Assessment
  • Spatial Analysis
  • Geographically Weighted Regression
  • AVM
  • CAMA

Cite this

@article{eb5d8571b651467583fff6a753c93218,
title = "Accounting for Locational, Temporal, and Physical Similarity of Residential Sales in Mass Appraisal Modeling: Introducing the Development and Application of Geographically, Temporally, and Characteristically Weighted Regression (GTCWR)",
abstract = "Geographically weighted regression (GWR) has been recognized in the assessment community as a viable automated valuation model (AVM) to help overcome, at least in part, modeling hurdles associated with location, such as spatial heterogeneity and spatial autocorrelation of error terms. Although previous researchers have adjusted the GWR weights matrix to also weight by time of sale or by structural similarity of properties in AVMs, the research described in this paper is the first that has done so by all three dimensions (i.e., location, structural similarity, and time of sale) simultaneously. Using 24 years of single-family residential sales in Fairfax, Virginia, we created a new locally weighted regression (LWR) AVM called geographically, temporally, and characteristically weighted regression (GTCWR) and compared it with GWR-based models with fewer weighting dimensions.GTCWR was the only model to achieve IAAO-accepted levels of the coefficient of dispersion (COD), price-related differential (PRD), price-related bias (PRB), and median assessment-to-sale price ratio in both the training and testing samples, although it did not fully correct the existence of heteroscedasticity. With lower PRD and PRB levels, the application of temporal weighting to this data set did appear to help reduce indicators of vertical inequity. Along with an equitable, uniform, and defensible methodology that mirrors the sales comparison, GTCWR presents a new AVM that demonstrates an ability to value over 24 years of sales at IAAO standard levels, without the creation and implementation of time-based variables, the trimming of outliers, and time-intensive model specification and calibration",
keywords = "Property Tax Assessment, Spatial Analysis, Geographically Weighted Regression, AVM, CAMA",
author = "Paul Bidanset and Michael McCord and Lombard, {John A} and Peadar Davis and William McCluskey",
note = "Reference text: References Bidanset, P.E., and J.R. Lombard. 2014a. “Evaluating Spatial Model Accuracy in Mass Real Estate Appraisal: A Comparison of Geographically Weighted Regression and the Spatial Lag Model.” Cityscape: A Journal of Policy Development and Research 16(3): 169-182. Bidanset, P.E., and J.R. Lombard. 2014b. “The Effect of Kernel and Bandwidth Specification in Geographically Weighted Regression Models on the Accuracy and Uniformity of Mass Real Estate Appraisal.” Journal of Property Tax Assessment & Administration 11(3): 5–14. Bidanset, P., and J.R. Lombard. 2017. “Optimal Kernel and Bandwidth Specifications for Geographically Weighted Regression: An Evaluation Using Automated Valuation Models (AVMs) for Mass Real Estate Appraisal.” In Applied Spatial Modelling and Planning, edited by J. Lombard, E. Stern, and G. Clarke, 107–120. Abingdon, United Kingdom: Routledge. Borst, R.A. 2014. “Improving Mass Appraisal Valuation Models Using Spatio-Temporal Methods.” International Property Tax Institute, Toronto, Ontario, Canada. Borst, R., and W. McCluskey. 2008. “Using Geographically Weighted Regression to Detect Housing Submarkets: Modeling Large-Scale Spatial Variations in Value.” Journal of Property Tax Assessment & Administration 5(1): 21–54. Brunsdon, C. 1998. “Geographically Weighted Regression: A Natural Evolution of the Expansion Method for Spatial Data Analysis.” Environment and Planning A(30): 905–1927. Brunsdon, C., A.S. Fotheringham, and M.E. Charlton. 1996. “Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity.” Geographical Analysis 28: 281–298. Crespo, R., A.S. Fotheringham, and M. Charlton. 2007. “Application of Geographically Weighted Regression to a 19-Year Set of House Price Data in London to Calibrate Local Hedonic Price Models.” In Proceedings of the 9th International Conference on GeoComputation, edited by U. Demšar. Maynooth, Ireland: NCG, NUI. Fannie Mae. 2017. “Selling Guide: B4-1.3-08: Comparable Sales (1/31/17),” April 26. https://www.fanniemae.com/content/guide/selling/b4/1.3/08.html (accessed May 2017). Fotheringham, A.S., C. Brunsdon, and M. Charlton. 2002. Geographically Weighted Regression: the Analysis of Spatially Varying Relationships. Chichester, England: Wiley. Fotheringham, A.S., R. Crespo, and J. Yao. 2015. “Geographical and Temporal Weighted Regression (GTWR).” Geographical Analysis 47(4): 431–452. Gloudemans, R.J., and R.R. Almy. 2011. Fundamentals of Mass Appraisal. Kansas City, Mo.: International Association of Assessing Officers. Gollini, I., B.M. Charlton, C. Brunsdon, and P. Harris. 2013. “GWmodel: An R Package for Exploring Spatial Heterogeneity using Geographically Weighted Models.” arXiv preprint arXiv:1306.0413. Guo, L., Ma, Z., & Zhang, L. 2008. Comparison of bandwidth selection in application of geographically weighted regression: a case study, 38(9), 2526-2534. Huang, B., B. Wu, and M. Barry. 2010. “Geographically and Temporally Weighted Regression for Modeling Spatio-Temporal Variation in House Prices.” International Journal of Geographical Information Science 24(3): 383–401. IAAO. 2003. “Standard on Automated Valuation Models (AVMs).” Assessment Journal 10(4): 109–154. Jiang, R., H. Wang, B. Huang, and G. Guo. 2013. “An Improved Geographically and Temporally Weighted Regression Model with a Novel Weight Matrix.” In Proceedings of the 12th International Conference on GeoComputation. Wuhan, China. Lockwood, T., and P. Rossini, P. 2011. “Efficacy in Modelling Location within the Mass Appraisal Process.” Pacific Rim Property Research Journal 17(3): 418–442. McCluskey, W.J., M. McCord, P.T. Davis, M. Haran, and D. McIlhatton. 2013. “Prediction Accuracy in Mass Appraisal: A Comparison of Modern Approaches.” Journal of Property Research 30(4): 239–265. McMillen, D.P. 1996. “One Hundred Fifty Years of Land Values in Chicago: A Nonparametric Approach.” Journal of Urban Economics 40(1): 100–124. McMillen, D.P., and C.L. Redfearn. 2010. “Estimation and Hypothesis Testing for Nonparametric Hedonic House Price Functions.” Journal of Regional Science 50(3): 712–733. Moore, J.W. 2009. “A History of Appraisal Theory and Practice: Looking back from IAAO’s 75th Year.” Journal of Property Tax Assessment & Administration 6(3): 23–50. Moore, J.W., and J. Myers. 2010. “Using Geographic-Attribute Weighted Regression for CAMA Modelling.” Journal of Property Tax Assessment & Administration 7(3): 5–28. Shi, H., L. Zhang, and J. Liu, J. 2006. “A New Spatial-Attribute Weighting Function for Geographically Weighted Regression.” Canadian Journal of Forest Research 36(4): 996–1005.",
year = "2018",
month = "1",
day = "31",
language = "English",
volume = "14",
pages = "5--13",
number = "2",

}

TY - JOUR

T1 - Accounting for Locational, Temporal, and Physical Similarity of Residential Sales in Mass Appraisal Modeling: Introducing the Development and Application of Geographically, Temporally, and Characteristically Weighted Regression (GTCWR)

AU - Bidanset, Paul

AU - McCord, Michael

AU - Lombard, John A

AU - Davis, Peadar

AU - McCluskey, William

N1 - Reference text: References Bidanset, P.E., and J.R. Lombard. 2014a. “Evaluating Spatial Model Accuracy in Mass Real Estate Appraisal: A Comparison of Geographically Weighted Regression and the Spatial Lag Model.” Cityscape: A Journal of Policy Development and Research 16(3): 169-182. Bidanset, P.E., and J.R. Lombard. 2014b. “The Effect of Kernel and Bandwidth Specification in Geographically Weighted Regression Models on the Accuracy and Uniformity of Mass Real Estate Appraisal.” Journal of Property Tax Assessment & Administration 11(3): 5–14. Bidanset, P., and J.R. Lombard. 2017. “Optimal Kernel and Bandwidth Specifications for Geographically Weighted Regression: An Evaluation Using Automated Valuation Models (AVMs) for Mass Real Estate Appraisal.” In Applied Spatial Modelling and Planning, edited by J. Lombard, E. Stern, and G. Clarke, 107–120. Abingdon, United Kingdom: Routledge. Borst, R.A. 2014. “Improving Mass Appraisal Valuation Models Using Spatio-Temporal Methods.” International Property Tax Institute, Toronto, Ontario, Canada. Borst, R., and W. McCluskey. 2008. “Using Geographically Weighted Regression to Detect Housing Submarkets: Modeling Large-Scale Spatial Variations in Value.” Journal of Property Tax Assessment & Administration 5(1): 21–54. Brunsdon, C. 1998. “Geographically Weighted Regression: A Natural Evolution of the Expansion Method for Spatial Data Analysis.” Environment and Planning A(30): 905–1927. Brunsdon, C., A.S. Fotheringham, and M.E. Charlton. 1996. “Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity.” Geographical Analysis 28: 281–298. Crespo, R., A.S. Fotheringham, and M. Charlton. 2007. “Application of Geographically Weighted Regression to a 19-Year Set of House Price Data in London to Calibrate Local Hedonic Price Models.” In Proceedings of the 9th International Conference on GeoComputation, edited by U. Demšar. Maynooth, Ireland: NCG, NUI. Fannie Mae. 2017. “Selling Guide: B4-1.3-08: Comparable Sales (1/31/17),” April 26. https://www.fanniemae.com/content/guide/selling/b4/1.3/08.html (accessed May 2017). Fotheringham, A.S., C. Brunsdon, and M. Charlton. 2002. Geographically Weighted Regression: the Analysis of Spatially Varying Relationships. Chichester, England: Wiley. Fotheringham, A.S., R. Crespo, and J. Yao. 2015. “Geographical and Temporal Weighted Regression (GTWR).” Geographical Analysis 47(4): 431–452. Gloudemans, R.J., and R.R. Almy. 2011. Fundamentals of Mass Appraisal. Kansas City, Mo.: International Association of Assessing Officers. Gollini, I., B.M. Charlton, C. Brunsdon, and P. Harris. 2013. “GWmodel: An R Package for Exploring Spatial Heterogeneity using Geographically Weighted Models.” arXiv preprint arXiv:1306.0413. Guo, L., Ma, Z., & Zhang, L. 2008. Comparison of bandwidth selection in application of geographically weighted regression: a case study, 38(9), 2526-2534. Huang, B., B. Wu, and M. Barry. 2010. “Geographically and Temporally Weighted Regression for Modeling Spatio-Temporal Variation in House Prices.” International Journal of Geographical Information Science 24(3): 383–401. IAAO. 2003. “Standard on Automated Valuation Models (AVMs).” Assessment Journal 10(4): 109–154. Jiang, R., H. Wang, B. Huang, and G. Guo. 2013. “An Improved Geographically and Temporally Weighted Regression Model with a Novel Weight Matrix.” In Proceedings of the 12th International Conference on GeoComputation. Wuhan, China. Lockwood, T., and P. Rossini, P. 2011. “Efficacy in Modelling Location within the Mass Appraisal Process.” Pacific Rim Property Research Journal 17(3): 418–442. McCluskey, W.J., M. McCord, P.T. Davis, M. Haran, and D. McIlhatton. 2013. “Prediction Accuracy in Mass Appraisal: A Comparison of Modern Approaches.” Journal of Property Research 30(4): 239–265. McMillen, D.P. 1996. “One Hundred Fifty Years of Land Values in Chicago: A Nonparametric Approach.” Journal of Urban Economics 40(1): 100–124. McMillen, D.P., and C.L. Redfearn. 2010. “Estimation and Hypothesis Testing for Nonparametric Hedonic House Price Functions.” Journal of Regional Science 50(3): 712–733. Moore, J.W. 2009. “A History of Appraisal Theory and Practice: Looking back from IAAO’s 75th Year.” Journal of Property Tax Assessment & Administration 6(3): 23–50. Moore, J.W., and J. Myers. 2010. “Using Geographic-Attribute Weighted Regression for CAMA Modelling.” Journal of Property Tax Assessment & Administration 7(3): 5–28. Shi, H., L. Zhang, and J. Liu, J. 2006. “A New Spatial-Attribute Weighting Function for Geographically Weighted Regression.” Canadian Journal of Forest Research 36(4): 996–1005.

PY - 2018/1/31

Y1 - 2018/1/31

N2 - Geographically weighted regression (GWR) has been recognized in the assessment community as a viable automated valuation model (AVM) to help overcome, at least in part, modeling hurdles associated with location, such as spatial heterogeneity and spatial autocorrelation of error terms. Although previous researchers have adjusted the GWR weights matrix to also weight by time of sale or by structural similarity of properties in AVMs, the research described in this paper is the first that has done so by all three dimensions (i.e., location, structural similarity, and time of sale) simultaneously. Using 24 years of single-family residential sales in Fairfax, Virginia, we created a new locally weighted regression (LWR) AVM called geographically, temporally, and characteristically weighted regression (GTCWR) and compared it with GWR-based models with fewer weighting dimensions.GTCWR was the only model to achieve IAAO-accepted levels of the coefficient of dispersion (COD), price-related differential (PRD), price-related bias (PRB), and median assessment-to-sale price ratio in both the training and testing samples, although it did not fully correct the existence of heteroscedasticity. With lower PRD and PRB levels, the application of temporal weighting to this data set did appear to help reduce indicators of vertical inequity. Along with an equitable, uniform, and defensible methodology that mirrors the sales comparison, GTCWR presents a new AVM that demonstrates an ability to value over 24 years of sales at IAAO standard levels, without the creation and implementation of time-based variables, the trimming of outliers, and time-intensive model specification and calibration

AB - Geographically weighted regression (GWR) has been recognized in the assessment community as a viable automated valuation model (AVM) to help overcome, at least in part, modeling hurdles associated with location, such as spatial heterogeneity and spatial autocorrelation of error terms. Although previous researchers have adjusted the GWR weights matrix to also weight by time of sale or by structural similarity of properties in AVMs, the research described in this paper is the first that has done so by all three dimensions (i.e., location, structural similarity, and time of sale) simultaneously. Using 24 years of single-family residential sales in Fairfax, Virginia, we created a new locally weighted regression (LWR) AVM called geographically, temporally, and characteristically weighted regression (GTCWR) and compared it with GWR-based models with fewer weighting dimensions.GTCWR was the only model to achieve IAAO-accepted levels of the coefficient of dispersion (COD), price-related differential (PRD), price-related bias (PRB), and median assessment-to-sale price ratio in both the training and testing samples, although it did not fully correct the existence of heteroscedasticity. With lower PRD and PRB levels, the application of temporal weighting to this data set did appear to help reduce indicators of vertical inequity. Along with an equitable, uniform, and defensible methodology that mirrors the sales comparison, GTCWR presents a new AVM that demonstrates an ability to value over 24 years of sales at IAAO standard levels, without the creation and implementation of time-based variables, the trimming of outliers, and time-intensive model specification and calibration

KW - Property Tax Assessment

KW - Spatial Analysis

KW - Geographically Weighted Regression

KW - AVM

KW - CAMA

M3 - Article

VL - 14

SP - 5

EP - 13

IS - 2

ER -