Akademik Veri Yönetim Sistemi
LAND USE POLICY, cilt.167, ss.108027, 2026 (Scopus)
The factors influencing property values have been studied across a wide range of variables, methods, and regions. However, studies that examine property prices using a wide range of current methods and 18 variables across dimensions such as environmental, physical, economic, migration, and transportation are limited. The primary objective of this study is to examine the effects of these variables on property prices. In addition to variables frequently used in the literature, eight new variables, such as bus and minibus line length, air quality values, average elevation, resilience score, and Syrian refugee density, were evaluated comparatively using current methods such as Ordinary Least Squares (OLS), Multiscale Geographically Weighted Regression (MGWR), Geospatial Random Forest (GeoRF), and SHAP. According to the MGWR results: 1) Property prices are negatively associated with building age, number of floors, population density, LST, PM10, and proximity to major roads and rail stations. 2) Property prices are positively associated with NO₂, floor area ratio, socioeconomic score, average building height, and hospital accessibility. According to the SHAP results generated from the GeoRF results: 1) The top five variables most influential in estimating property prices are the square meter value, the socioeconomic score, the age of the building, NO₂, and LST. 2) The variable with the most negative impact on the average property price is the LST, while the variable with the most positive impact is the NO₂ level. The research findings are expected to be useful in developing spatial strategies for citizens, researchers, policymakers, and real estate market professionals.