Survey experiments have become central in the social science methods toolbox, promising the dual benefit of an experimental design (for causal inference) on population-representative samples (for external validity). In addition, conjoint survey experiments promise a causal understanding of the complex, multidimensional choices that citizens and politicians regularly face in political decision-making - like voting for parties or deciding between policy alternatives. This course provides a short introduction to this survey experimental method, outlining the basic logic, summarizing current topical debates on promises and pitfalls, and giving hands-on advice on design and implementation. We expect and encourage you to bring your own conceptual ideas for a survey-experimental research question that you want to develop further during the course.

 

We alternate between introductory lectures, discussion sessions and group work. Discussion focuses on the practical set-up of conjoints (see detailed description in the day-to-day schedule of the syllabus), also allowing you to raise aspects of your own research projects and discuss with others who plan similar designs.