Hedonic price models with geographically weighted regression: An application to hospitality

Abstract

Abstract The objective of this study was to propose and test a methodology that allow destination managers and hoteliers improve the allocation of resources. For this purpose, this paper analysed the impact of both establishment (e.g. category, size and location) and assessment variables of services included in hotel room prices using hedonic price regression and geographically weighted regression (GWR). The data were collected in the low season using TripAdvisor and Google Maps for 57 hotels located in Malaga. Analyses showed that spatial correlation creates different patterns of quality-value perceptions within the same city, which is an advance in the knowledge about the hotel location decision–making processes and their implications on destination marketing. These competitive subsystems cannot be detected with the use of ordinary least squares alone. Although the values extracted using a hedonic price model are consistent with the previous literature, the presence of geographic variability in the estimated hedonic model coefficients might be misleading for some hotels. The fitting coefficient of the GWR confirms the need to incorporate GWR into hedonic price models.