Causality Does Influence Conjunctive Probability Judgments If Context and Design Allow for It

In this paper we study and discuss factors that may affect the frequency of the conjunction fallacy, in particular the role of perceived causality. A previous study by Thuring and Jungermann (1990) failed to establish a relation between causality and the conjunction fallacy, a result incongruent with the original analysis by Tversky and Kahneman (1983). We doubt the generalizability of this result, and propose that it was due to the specific experimental conditions (between-subjects manipulation of causality). Under different conditions (within-subject manipulation) an effect of perceived causality may well be observed. In two experiments, significant effects of causality were found. These and previous findings are discussed and put into perspective.