Log in / Register
Home arrow Economics arrow Bounded rational choice behaviour
< Prev   CONTENTS   Next >

5.4. Conclusion and Future Work

Increasing evidence suggests that choice behaviour in real-world settings may be more guided by principles of bounded rationality as opposed to typically assumed fully rational, utility-maximizing principle. Moreover, many choices and decisions in real-world settings are context-dependent. Thus, people may simplify the decision problem by considering information selectively, limiting choice sets, using simplifying decision heuristics, etc. Under such circumstances, conventional rational choice models do not validly mimic the decision processes. To model both outcomes and processes of decision-making, this chapter proposed a modeling approach which incorporates attribute thresholds as the basic mechanism to model attribute representation. It has been shown that such a discrete cognitive representation has the generality to be the source of several decision heuristics, including the typical conjunctive, disjunctive and lexicographic rules, assuming that the overall judgment threshold varies with individual and contextual differences. This approach allows modeling unobservable decision heterogeneity involved in a single decision, for example, in the form of a latent-class specification. Furthermore, the choice of decision strategies is modeled, taking into account mental effort, risk perception and expected outcome as explanatory factors. It should be emphasized that in this case the logit form represents the functional relationship between probabilistic choices and input variables and is not derived from assumptions of utility-maximizing behaviour.

Two types of decision models are derived under the heterogeneous heuristic modeling framework: one models the satisficing decision; the other models the comparative decision. Both models are applied to a decision problem pertaining to pedestrian shopping behaviour and compared with conventional MNL. The results show that the proposed models may not be necessarily superior to conventional logit models in terms of model selection criteria due to the extra complexity related to modeling preference structure and heuristic selection. However, the new models suggest more interesting insights in the underlying decision processes. Inference can be made probabilistically whether a factor is used for the decision, the sequence of factor processing, the effort involved, the risk propensity and the outcome preference of the decision-maker, which together provide much richer information than what conventional utility-maximizing models can offer.

Understanding decision processes additional to outcomes is a promising research direction. A more developed model should take into account more contextual and socio-demographic factors in the heuristic selection part, which then can provide possibilities to answer the question of who making decisions in what context and how. Of course, these assumptions of information processing must be subject to empirical tests to validate the model. Some tests need carefully designed experiments to test the ability of the model to reveal the specific decision style under control. Objective measures such as mental effort and risk propensity should be devised as criteria for validating the proposed elements in the heuristic choice part.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science