The standard implementation of hedonic modelling techniques is non-spatial, therefore the complicating effects of spatial correlation are typically ignored. The latest applications of hedonic dwelling price models have included recent advances in spatial analysis that control for spatial dependence and heterogeneity. The study of spatial aspects of hedonic modelling pertains to spatial econometrics which is relevant to this study because it clearly accounts for the influence and peculiarities caused by space in real estate price modelling analysis.
The main goal of this work is to provide location sensitive hedonic price model as statistically justifiable means of reviewing the values and valuation with the model specification capturing adequately the significance and proper consideration of spatial elements. Besides the regression method, which is the most widely used method in obtaining econometric models, and lately used geographically weighted regression, in this research we introduced regression kriging, geostatistical method not yet used in this manner in real estate analysis. The geostatistical methods are characterized by the fact that they use the spatial structure of correlation to explain the dwelling price.
Given that a cross-sectional analysis of house prices involves georeferenced information, Geographical Information System (GIS) and spatial statistics are suitable tools for hedonic modelling. Recently developed R language packages designed for managing, processing and visualisation data given in GIS formats facilitate the advance approach in this field. The spatial predictors, given as raster maps, were used as auxiliary inputs necessary for regression modelling. In addition to standard environmental predictors, some socio-economic data such as distribution, ages and income of inhabitants, were prepared in the same manner enabling their use in GIS support environment. The neighborhood characteristics and socio-economic variables for each household were not taken into consideration in the investigated model.
A spatial-econometric hedonic dwelling price model is developed and estimated for the Belgrade metropolitan area based on cross-sectional and georeferenced transaction data. Confirmatory spatial data driven approach was used in spatial hedonic modelling analysis and the continuous price maps reflecting patterns in the spatial distribution of location price within the city were drawn up.