Abstract
A new perspective is available for the relationship between transportation systems and property prices, “evaluation by considering bandwidths”. The term “bandwidth” indicates the number of neighbor data points having an influence on the target point. The main motivation of this research is eliminating the restricted effects of a spatial analysis. When a single bandwidth value is defined for all models, indicate that an equal number of neighbor data in GWR influences all factors of the model for target points. When the multiscale geographically weighted regression (MGWR) is used, a unique bandwidth value can be defined for each factor. This study, by using MGWR, provides researchers the ability of distinguishing the effects of factors and classifying them as either wide-range or narrow-range. A total of 3487 geographical data were collected for a study area covering Esenyurt and Beylikduzu Counties in Istanbul, Turkey. As a result, when considering different bandwidths, factors are indeed recognized as either narrow-range or wide-range. The output of this study proved that the researchers are strongly recommended considering the bandwidth effects while performing a spatial regression analysis. Hence, the gap in the selection of factors for spatial analyses in different study areas is likely to be fulfilled.