Discrete Choice Modelling

Aim

After completion of the course, the student should on an advanced level be able to:

• Identify problems most suitably modeled with discrete choice methods;
• Describe which models are suitable for specific applications using panel or cross-sectional data;
• Develop appropriate discrete choice model specifications;
• Create appropriate data structures for estimating discrete choice models;
• Critically review and interpret model results of statistically complex discrete choice models;
• Use statistical software to estimate discrete choice models, calculate predictions, and interpret and analyze results.

Content

This course enables students to perform their own empirical research using discrete choice methods. Students learn how to create discrete choice datasets, estimate discrete choice models, including binomial, multinomial, and conditional logistic regression, and interpret model output. The focus will be on the practical aspects of modeling. During intensive computer labs, hands on experience will be provided using real data drawn from examples in the areas of consumer choice, migration, and labor market mobility. More advanced models for handling panel data and unobservable heterogeneity, as well as identification of latent groups will be examined and deployed. Applications to counterfactual and agent-based simulation will also be explored during lab sessions.