Goal-based models for discrete choice analysis

- Goals direct decision making, from the most abstract levels of motivation to the details of evaluation of options.
- We formulate goal-based choice models that explain the impacts of the decision context on the decision maker.
- We extend standard utility maximization models to multiple and simultaneous goal pursuit.
- We develop new goal-based models for empirical testing in transportation.
- We suggest a possible integrative goal-based representation.

Goals direct decision making, from the most abstract levels of motivation to the multitudinous details of evaluation of options available for choice. However, the pervasive influence of goals in decision processes is generally not explicitly recognized at the level of demand model formulation and specification. In applied economics generally, and transportation specifically, demand models relate product/service attributes directly to behavior, using utility (or value) as a shorthand representation for the impact of goals. In this paper we argue that this is a limiting view that restricts our thinking about decision making and, hence, our representation and inference-making about that behavior. We support this argument by reinterpreting and/or extending various applications of hybrid models in transportation to a goal-based framework and formulating goal-based choice models which recognize that goals (1) are drivers of choice, (2) explain the choice of strategy, (3) are part of the constraint set and (4) contribute to explaining impacts of the decision context on the allocation of cognitive resources by the decision maker.