FUNCTIONAL ANALYSIS OF MODE CHOICE

This paper develops a relatively new paradigm for mode-choice behavior modeling. The paradigm emphasizes functional form by establishing functions that relate subjective evaluations of transportation system attributes to objective levels of these attributes and functions that relate observed mode choices and preferences to combinations of subjective impressions. These functions are derived from a theory of decision making and behavior that views the decision maker as an integrator of information. According to this theory, the overall evaluation of a given transportation system can be represented as an algebraic combination of the traveler's evaluations of the various attributes of the system. Two experiments were conducted to evaluate this paradigm. Questionnaires were distributed to respondents, who were asked to indicate their degree of preference for car or bus for each of a series of hypothetical mode-choice situations. These situations were generated by combining varying levels of time difference favoring car over bus, cost difference favoring bus over car, and number of riders in the car. Each judgment thus required a trade-off of cost, time, and interpersonal factors. Cluster analysis was used to separate respondents into distinct subgroups of homogeneous decision makers. These subgroups differed in terms of overall preference for car or bus and the relative weighting or trade-off of factors. Actual mode choice for work trips was then predicted on the basis of preference responses to the hypothetical mode-choice situations, estimates of cost and time factors for individual respondents, and transportation availability constraints. A high level of predictive validity was attained in each experiment. It is suggested that the present paradigm may be useful for analyzing traveler decision processes, for estimating latent demand for alternative transportation opportunities, and for predicting responses to altered or new transportation systems. /Authors/