Multi-modal mobility decisions: A consumer's perspective
We will develop new models of consumer choice that go beyond traditional choice models in the social and economic sciences, which focus on the selection of one option from many (i.e., one transportation mode per trip), to modeling the choice of a subset of mobility options for a single trip (e.g., drive a car to a park & ride, then use public transportation for the remainder of the commute). The project will make use of existing data and models as well as small, targeted data collection needed to identify the parameters of the new models, test the key assumptions of the new models, and validate predictions that emerge from our initial modeling exercises.
The targeted data collection will then feed into a revised modeling tool that can be used to identify business opportunities in the domain of multi-modal mobility. We will use and extend a variety of modeling methods and approaches including discrete choice models (e.g., logistic and probit models that model both systematic and unobserved portions utility, consideration set models where choices are made from a subset of modes pre-selected by the consumers), behavioral economic models (expected utility models and their extensions), behavioral decision making models (e.g., the Nobel prize winning prospect theory, which incorporates, say, risk-aversion tendency and loss aversion in the decision), monte carlo simulation, "what if" scenario modeling, and state-of-the-art statistical modeling tools. The research program will facilitate understanding the factors that promote and inhibit uptake by consumers of multi-modal mobility options.
The business potential of those factors will be based on state-of-the-art research in behavioral decision making, will make use of both existing data and new data collected for the project, and will be flexible to permit other uses such as planning and forecasting. The team will leverage existing connections in US cities (e.g., Portland, Los Angeles) as well as global regions (e.g., major cities in China, including Beijing and Shanghai). The project will include testing the predictive performance of these models against real-world outcomes in selected cities.
- Richard D. Gonzalez
2013-09-01 - 2015-12-31