Selecting independent variables in the hedonic analysis of hotel rooms

Abstract

The hedonic pricing method has been applied in the analysis of hotel room rates since the early 1990s and ever since has become the norm in the investigation of hotel room price determinants. A usual problem in the specific literature is that there are no theoretical guidelines in selecting the explanatory variables in hedonic pricing models. Therefore, scholars employed in such kind of research face the challenge of which of the numerous price determinants proposed in the literature should they include in their own analysis, with the hazard of multicollinearity complicating the issue even further. The current article proposes a strategy where both the variable selection problem and the potential multicollinearity are faced by utilizing a combination of stepwise regression techniques with a proper categorization of the candidate regressors. This procedure is used in analysing the determining factors of the hotel accommodation prices in Cyprus, a country whose economy is highly dependent on tourism, for 12 consecutive months, in order to account for the inherent seasonality. The main findings suggest that adding some attributes to the advertisements of the hotels significantly increase hoteliers’ remunerations, but advertising certain other features may be actually sending negative signals, therefore hoteliers should consider excluding them from their brochures.