University of California Transportation Center

With the introduction and increasing reliance on the activity based approach in travel demand analysis and forecasting, discrete choice methods are more often in spatial contexts (e.g., residential location, job location, destination/activity choice). A necessity in specifying spatial choice models is the inclusion of the alternatives decision makers consider, and a realistic inclusion in the specification of the attributes of these alternatives, the characteristics of the decision making context, and the relevant characteristics of the decision maker. These details describe differences that exist among choices and individuals making choices. In the case of travel behavior, attributes of the alternatives have traditionally included attributes such as cost, distance, time, level of service and opportunities. Researchers however have recognized the benefit of attitudes in the estimation of choice models, showing improvement in explanatory power with the inclusion of attitudinal attributes of the individual. There is still however a vast expanse of unexplored attitudinal attributes in choice modeling. Particularly lacking in choice modeling is a strong theoretical underpinning of attitude formation and attitude relation to choice. Theorists in geography in the mid to late 1970s recognized and developed one such theory regarding the emotional and attitudinal association that people have with places. This became known as the theory of sense of place, which is the “affective ties with the material environment” (Tuan, 1974). This theory presents great potential in furthering the descriptive power of choice models, particularly destination choice. However, challenges abound, as this theory is rich in development, but poor in computational operationalization. In addition, everyday activity locations have not been adequately explored in sense of place quantification. In this paper, an overview of developments in discrete choice methods is presented, followed by a discussion of sense of place. Current and future work is discussed and benefits to choice modeling are presented.