An Evaluation of the Role of Inductive Confirmation in Relation to the Conjunction Fallacy

Abstract

Inductive confirmation has been proposed as a mechanism giving rise to the conjunction fallacy. For each of five separate vignettes, probability estimates were obtained for a neutral event, for a second event: i.e. the “added conjunct”, and for their conjunction. The added conjunct was selected such that it was inductively confirmed, either by some background evidence provided in the vignette or by the other component event. So as to achieve sufficient statistical power, multilevel models were used to analyse the data. For the added conjunct, the level of confirmation and the posterior probability were significantly associated such that higher levels of confirmation were associated with larger probability estimates. However, there was no significant association between the level of confirmation on the one hand and the incidence of the fallacy and the conjunctive probability on the other.