Im Rahmen des Forschungsseminars laden wir recht herzlich ein zum Vortrag:
Dr. Ralf Mayrhofer (Univ. Göttingen)
Violations of Markov causality in human causal reasoning
(with a preface by Prof. Dr. Gerhard Schurz)
Abstract
In psychology, it is quite popular to model human causal reasoning with causal Bayesian networks, a central assumption of which is the Markov constraint. According to the Markov constraint, an inference from a cause X to its effect Y should be invariant across conditions in which causes or other effects of X are present or absent. However, a decade of psychological research has shown that human causal reasoning does not adhere to the Markov constraint as inferences typically depend upon variables that the inference should be independent of. In this talk, I will present results of experimental studies and several psychological explanations for this phenomenon.
Speaker
Ralf Mayrhofer is a postdoctoral researcher at the Institute of Psychology, University of Göttingen; his DFG-funded research project is part of the priority program “New frameworks of rationality” (DFG-SPP 1516). He holds a PhD in experimental psychology and statistics and published several papers on human causal learning and reasoning in journals such as Psychological Review, Cognition, and Cognitive Science. Besides his interests in human causal reasoning, he is also interested in the human aspects of categorization, explanation and the perception of simple physical interactions (e.g., the collision of Billiard balls) as well as, more generally, Bayesian modeling of human cognition.