I have several interesting Master's projects to offer. My main interests are decision making in risky situations and the relation between response speed and choice. The project on the vorlesungsverzeignis is only one direction in which you could go. Below, I list ideas for projects that you might like to write your master's thesis about. If you are interested in one or more of these projects, I can send you some additional introduction to the projects and relevant literature.

Factors that Influence how People Perceive Risk and Choose in a Risky Choice Situation

In everyday life, you face many decisions: Do I go left or right, which hairdresser do I choose, which bike do I buy? In some cases, you know on beforehand the exact results of the various choice options at hand. However, the vast majority of decisions have to be made in uncertainty about the possible outcomes. Still, you probably make many of these uncertain, risky decisions with reasonable confidence. In this master's project, we aim to get a deeper understanding of how you make these risky decisions by studying factors that influence people's choice behavior, such as outcome uncertainty, information cost, and speed stress.

In the first semester, we review the literature to learn what we know about factors that influence risky decision making. In the second semester, the research goals and hypotheses for individual students' projects will be worked out. In the third semester, the goals and hypothesis will be implemented in an empirical study.

Validation of the Diffusion Model

The diffusion model is a very succesfull model for two-choice response time data. More and more studies apply the diffusion model these days to analyse response time data. Response time data generally contains two variable: response time and accuracy (or choice). The problem of these bivariate data lies in the fact that those two variables are in trade-off: when a participants tries to speed up, her accuracy will generally go down; when a participant wants to become more accurate, she often has to slow down to do so. The most important strength of the diffusion model is that it allows to translate these complicated data into more easily interpretable model parameters. These model parameters are assumed to reflect underlying psychological components of decision making, such as the speed of information processing or the caution with which a participant responds. So, the model allows to learn more about the data than would be possible from the analysis of response time and accuracy separately.

Although the underlying mathematical theory is sound, the validity of the translation to psychological components of decision making naturally has to be backed up by empirical validation studies. In such a validation study, an experimental manipulation of, e.g., response caution ("be cautious!") should result in specific effects on the related model parameter. Although some validation studies have been performed, the vast number of studies that apply the diffusion model today asks for a more regorous and comprehensive test of interpretational validity of the model. A master's project on this subject would have a very good chance to make a serious contribution to the field of response time modeling.

One Process or Separate Stages of Processing?

Most models for response time implicitly assume that the trade-off between speed and accuracy is a continuous function: a participant who is performing highly accurately can speed up more and more at the cost of getting less and less accurate. In our 2010 paper A phase transition model for the speed-accuracy trade-off in response time experiments., we showed that this assumption of continuity might not hold if participants are pressed to the guessing extreme of the speed-accuracy curve. An interesting master's project would be to design an experiment that offers some additional evidence for this discontinuous, two-stage account of information processing in response time tasks.
gilles curriculum vitae