Course Announcement: Choice Modeling
I would like to announce a course on Choice Modeling resp. Choice-Based Conjoint Analysis given at the ECPR-Summer School in Ljubljana / Slowenia from 30th Jul. to 14th Aug. 2010. The course is of interest not only for business economics and applied social research, but also for simulation modelers since it deals with the empirical micro-level calibration of decision functions of possibly heterogeneous agents.
Also, the course instructor is a scholar proficient in multi-agent simulation. Here are some course details, wich can also be found at
http://www.ecprnet.eu/summerschools/Ljubljana/course_details.asp?courseID=15
COURSE INSTRUCTOR
First Name: Gero
Last Name: Schwenk
Email: Gero.Schwenk[att]sowi.uni-giessen.de
<mailto:[log in to unmask]>
Institution: Justus-Liebig-Universität Giessen
COURSE DETAILS
Course Title: Discrete Choice Models for Political Behavior
Course Type: Main Course Full
Course Outline: PDF Not available
Virtually all human behavior is associated with choice – the most prominent examples include voting choice between different parties or candidates and consumer choice, but also travel mode choice and choice in the areas of education and employment. Analyis of choice settings and processes enables the modeler to reason about the causes of choices and therefore predict effects of policy-interventions. This makes choice modeling a very appealing tool for practical policy-planning in public adminsitration and business.
The course aims at providing it's participants with the knowledge necessary to both build and evaluate their own models and use them to asess plausible effects of policy-interventions by means of simulation.
Topics covered include modeling of limited depentent variables (Probit-, Logit-, Multinomial Logit- and Poisson Model), microeconomic and cognitive psychology theory of choice, design of choice experiments, design and implementation of policy-simulators in Excel, dependent choice alternatives (Nested Logit model) and modeling of preferences in heterogeneous populations (Mixed Logit Model). For estimation we will primarily use the LIMDEP and Nlogit software packages.
|