Metacognitive Monitoring in Wason Selection Task: The Influence of Content Abstractness, Conditional Type, and Social Content

In meta-reasoning studies, solvers are asked to make a metacognitive judgment about the correctness of their reasoning. It is often found that solvers have no insight into the actual accuracy of their responses and that they make confidence judgments based on cues such as the fluency of the response. According to the dual-process theory of reasoning, responses to reasoning tasks generated by heuristic processes (Type 1) are automatic, quick, and easy, while responses generated by analytic processes (Type 2) are often slower, require mental effort, and are deliberate. Participants tend to have more confidence in their Type 1 responses than in their type 2 responses, regardless of their accuracy. The Wason selection task involves a form of conditional reasoning. Typical results show that accuracy in the abstract version of the task is very low. This can be explained by cognitive biases such as confirmation and matching bias, which are based on heuristic processes. The aim of this study was to investigate how variations in content abstractness, conditional type, and social content affect judgments of confidence in Wason tasks. 128 participants took part in the 2x2x2 experiment. They solved variants of Wason tasks and rated their confidence after each response on a 0-100% scale. The results showed a statistically significant effect of social content (the participants were more confident in non-social content tasks than in social content tasks) and abstractness (more confidence in abstract content tasks than in concrete tasks). It appears that heuristically generated responses promote higher metacognitive scores. This result is consistent with the general views of the meta-reasoning framework and dual-process theories of reasoning.