Embedding Decision Heuristics in Discrete Choice Models: Assessing the MERITS of Majority of Confirming Dimensions, Extremeness Aversion, and Reference Revision
Year: 2013
Type: dissertation
Abstract: Contrary to the usual assumption of fixed, well-defined preferences, it is increasingly evident that individuals are likely to approach a choice task using decision heuristics that depend on the choice environment. These include heuristics defined by the local choice context, such as the gains or losses of an attribute value relative to the other attributes. Recent empirical findings also demonstrate that previous choices and previously encountered choice tasks can affect the current choice outcome, indicating a form of inter-dependence across choice sets. A number of these heuristics, namely the majority of confirming dimensions (MCD), the extremeness aversion and the reference revision heuristics, are analysed. These heuristics are not new, but their application, using the discrete choice modelling framework, to the transportation field has only barely begun. In particular, arising from the extremeness aversion heuristic, three models are discussed. The first is a recently developed model of context dependence known as the random regret minimisation (RRM) model. The second model is a non-linear utility model that makes reference to the worst attribute level in a choice set. The third model is a “relative advantage maximisation” (RAM) model, with an updated version of the existing RAM model introduced in this thesis. All these models are compared against one another and with the standard random utility maximisation (RUM) model. The results strongly indicate that incorporating context dependency into existing models should be a key consideration for the practitioner. Moreover, having identified some heuristics of special interest, the role of multiple heuristics in choice behaviour is also analysed. Interestingly, the heuristics themselves can be embedded directly into the utility functions by means of heuristic weighting functions, which weight the contribution of each heuristic to overall utility. The thesis examines the validity of such an approach.
Author: Wai Yie Leong
Institution: The University of Sydney
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W306516945
