공간계량분석방법을 이용한 부산 주택특성가격 모형 비교

• Published in: 한국비교정부학보 Vol. 15; no. 1; pp. 159 - 184
• Main Author: 정건섭(Kyoun-Sup Chung)
• Format: Journal Article
• Language: Korean
• Published: The Korean Association For Comparative Government 30.04.2011; 한국비교정부학회
• Subjects: 행정학; 공간적 자기상관; 공간계량모형; Hedonic Housing Prices; Geographically Weighted Regression; Spatial Econometrics Model; Spatial Autocorrelation; 주택특성가격; 지리적가중회귀모형; 공간가중행렬; Spatial Weight Matrix
• ISSN: 1598-964X; 2713-5357
• DOI: 10.18397/kcgr.2011.15.1.159

This study is trying to find out the spatial econometrics model in order to reflect the true reality in the housing price analysis. Due to the influence of spatial autocorrelation with respects to the housing price in the housing market, the estimation of the OLS is not preferable. Then the spatial econometrics model reflecting spatial dependence and spatial heterogeneity is the best alternative in terms of housing price estimation, probably. "Everything is related to everything else, but near things are more related than distant things". More realistic spatial weight matrix and spatial model scheme, better performance in the housing price prediction. Therefore, this study can show that the model improvement by spatial model scheme can be achieved by the comparing the spatial econometrics models such as SAR(Spatial Autoregressive Models), SEM(Spatial Errors Models), SAC(General Spatial Models), and GWR(Geographically Weighted Regression). Depends on the perspective of methodology, the SAC and the GWR models are preferable in terms of global(nomothetic approach) or local(idiographic approach) point of view, respectively. This study is trying to find out the spatial econometrics model in order to reflect the true reality in the housing price analysis. Due to the influence of spatial autocorrelation with respects to the housing price in the housing market, the estimation of the OLS is not preferable. Then the spatial econometrics model reflecting spatial dependence and spatial heterogeneity is the best alternative in terms of housing price estimation, probably. "Everything is related to everything else, but near things are more related than distant things". More realistic spatial weight matrix and spatial model scheme, better performance in the housing price prediction. Therefore, this study can show that the model improvement by spatial model scheme can be achieved by the comparing the spatial econometrics models such as SAR(Spatial Autoregressive Models), SEM(Spatial Errors Models), SAC(General Spatial Models), and GWR(Geographically Weighted Regression). Depends on the perspective of methodology, the SAC and the GWR models are preferable in terms of global(nomothetic approach) or local(idiographic approach) point of view, respectively. KCI Citation Count: 10