5. Practical Issues in the Specification and Estimation of Discrete Choice Models

The chapter focuses on the practical application of discrete choice models, particularly the multinomial logit model. It begins with an overview of the historical development of the logit model and its theoretical foundations. The core of the chapter is dedicated to the specification and estimation of utility functions, using a real-world dataset on transportation mode choices. Readers are guided through data preparation, the creation of utility functions, and the estimation of models using the mlogit package in R. The chapter also covers the interpretation of model results and the use of diagnostic tools to evaluate model performance. Throughout, the chapter emphasizes the importance of understanding the anatomy of utility functions and the practical skills required for effective model implementation.