Proceedings of the Annual Meeting of the Cognitive Science Society

Although young learners often express overconfidence, research has demonstrated that 3- to 4-year-old children maybe able to use a confidence scale to discriminate between their correct and incorrect responses. The current researchintroduces a novel paradigm to facilitate childrens reflection and reporting of confidence, based on the presentation ofdisconfirming evidence. This paradigm presents 3-, 4- and 5-year-olds with windows of varying occlusion (none, partial,and full). Children used a 3-point scale to assess their confidence that a target shape was behind each window. In fourconditions, we varied when and whether children received disconfirming evidence. Results suggest that violating childrensexpectations about the presence of the target shape on the first trial results in improves confidence calibration on futuretrials. Results also suggest that baseline confidence scale use improves with age. We discuss theoretical implications fordevelopment of uncertainty monitoring and potential applications of this novel paradigm.