The effects of source trustworthiness and inference type on human belief revision

Abstract

We investigated whether people revise their beliefs as a function of inference type or source trustworthiness. By doing so we aimed to find out if belief revision is better explained by mental model theory (Johnson-Laird & Byrne, 2002) or by a conditional probability view (Evans, Handley, & Over, 2003; Oaksford & Chater, 2001). We used modified modes ponens (MP) and modus tollens (MT) problems in which the first two premises were uttered by persons with varying degrees of trustworthiness. A third statement was presented as a fact and established inconsistency in the set of propositions. The participants' task was to indicate which of the first two premises they believed more after receiving the fact. We found that the belief in the conditional premise dropped significantly when this premise was stated by a low- rather than a high-trustworthy source. Moreover we found that the conditional premise was believed more in MT than in MP problems. Both findings are best explained by the conditional probability hypothesis (e.g., Evans et al., 2003).