Familiarity Effects and Questioning Biases in Human Belief Revision

Belief revision is the process by which one alters his or her belief state in the face of contradicting evidence. The contradictions typically arise from a set of statements in which not all propositions can be true at the same time. Despite widespread agreement that people have little difficulty finding such inconsistency, we still lack sufficient knowledge on how people revise their beliefs to resolve the inconsistencies. In the following paper, we report two experiments that were concerned with this research topic. In experiment 1, we explored how familiarity with the content of the statements (familiar vs unfamiliar) affects peoples’ belief revision choices. In experiment 2, we investigated whether different ways of questioning (what do you believe “more” vs. “less”) affect belief revision. The results show that both factors have a significant effect on peoples’ belief revision choices. Our results fit the predictions of the mismatch principle of the mental model theory of human reasoning (e.g. Johnson-Laird, Girotto, & Legrenzi, 2004).