Discrete Choice Analysis: Predicting Individual Behavior and Market Demand

Discrete choice models are widely used for the analysis of individual choice behavior and can be applied to choice problems in many fields such as economics, engineering, environmental management, urban planning, and transportation. For example, discrete choice modeling is used in marketing research to guide product positioning, pricing, product concept testing, and many other areas of strategic and tactical interest. Recent applications to predict changes in demand and market share include areas such as choice of travel mode, coffee brands, telephone service, soft drinks and other foods, financial services, internet access, and choice of durables such as smartphones, tablets, automobiles, air conditioners, and houses. This program also covers methods for online applications where predictions of individual choice behavior are used as inputs for the online optimization and personalization of advertising, recommendations and promotions. The methods covered include discrete choice models (logit, nested logit, generalized extreme value, probit, logit mixtures, hybrid choice models), data collection, specification, estimation, statistical testing, forecasting, and application. The covered topics include analysis of revealed and stated preferences data, sampling, simulation-based estimation, discrete panel data, Bayesian estimation, discrete-continuous models, menu choice, and integration of choice models with latent variables models. The course includes lab sessions where participants are provided with discrete choice software to learn how to use real databases to estimate and test discrete choice models taught in lectures and gain hands-on experience in using new discrete choice techniques for practical applications. By examining actual case studies of discrete choice methods, students will become familiar with problems of model formulation, estimation, testing, and forecasting.