Discrete Choice Modeling

PD Dr. Ossama Elshiewy

Target group:

  • Master students
    (MDM, UF, FRS, Wi-Inf, Wi-Päd, Steuerlehre, Wirtschafts & Sozialgeschichte, International Economics, Development Economics, Angewandte Statistik, Psychologie)

  • All PhD students (Faculty of Business and Economics) interested in discrete choice modeling

Learning goals:
  • Discrete choice modeling deals with analyzing choice behavior of individuals (e.g., consumers) as a function of variables that describe the choice alternatives (usually more than two alternatives).

  • After successful attendance the students will understand the methodological principles of discrete choice modeling.

  • Further, they will be able to estimate own discrete choice models using the statistical programming language R. (Previous knowledge in R is helpful!)

Course structure:
The course consists of two parts:
1. Attending a lecture (with integrated exercises)
2. Writing a term paper.

Important note: The lecture will be held in English, but the term paper can be written in either English or German.

Lecture content:
- Random Utility Theory
- Collecting Choice Data
---- Choice-based Conjoint
---- Consumer Purchase Data
- Analyzing Choice Data
---- Multinomial Logit (MNL) Models
---- Finite Mixture and Mixed MNL Models
---- Hierarchical Bayesian MNL Models

Term paper:
The term paper should contain a self-conducted empirical project.
Students will be provided with topics and empirical data, but are also welcome to analyze own projects.
Students are advised to use the statistical programming language R (and submit their code), but can be allowed to use different software in exceptional cases.

Train (2009). "Discrete Choice Methods with Simulation". 2nd Edition, Cambridge University Press.
Rossi, Allenby, and McCulloch (2005). "Bayesian Statistics and Marketing". Wiley.

Time and place:
The lecture will be offered again in the winter semester 2021/22.

Theoretical, methodological and empirical elaboration of a selected topic in discrete choice modeling.
Term paper (~6000 words)
Credits: 6 ECTS

Helpful prerequisites:
Basics in
- Probability theory and distributions
- Hypothesis testing
- (Logistic) Regression analysis