Modelling a decision maker's preferences: Testing for relevance and application development

The need for developing models in the tourism and hospitality industry which identify and make use of a decision maker's preferences has been demonstrated extensively in the works of Brown et al (2002) and Israeli and Brown (2004), as well as others (Ingold and Yeoman, 1997). In this study, the authors examine historical data of Israeli hotels in order to demonstrate empirically the need for these models and to determine which model best fits the empirical data. Three preference-based pricing models are tested, built with indifference curves to analyse the two-attribute case of room rate versus occupancy. These models differ from standard ones in that the decision maker's actual choices for specific budget scenarios are built directly into them. These constrained choice points are built into a table and are used to create an entire pricing schedule based on the preference ratio thus determined. The first model uses an unconstrained desired point and assumes that managers find rate and occupancy to be perfectly substitutable at all budget constraints. The second model assumes linear fractional substitutability and alters the ratio for specific constrained choice points (Israeli and Brown, 2004). The third model assumes perfect complementarity at each constrained choice point and assumes a consistent ratio for each constrained choice segment (Brown, 2004). The heuristic nature of the models makes statistical analysis of their validity difficult, but is thus attempted in any case. The motivation behind this test is to determine which model can best be used to evaluate behaviour and performance in a future study.