Article révisé par les pairs
Résumé : It is well established that anxiety influences a range of cognitive processes such as cognitive control or decision-making. What is less known is how anxiety influences the metacognitive evaluations individuals make about their own performance. The present study explored the importance of task-relatedness in the relation between anxiety and metacognitive awareness. More specifically, we hypothesized that induced anxiety would affect metacognitive performance in opposite directions depending solely on whether the threat is perceived as related or unrelated to the performance. We conducted two experiments to test this hypothesis. In both studies, electric shocks were administered randomly, regardless of participant's performance. In experiment 1, participants were informed that the shocks would be administered randomly, unrelated to performance. In experiment 2, participants were told that their metacognitive performance would influence the administration of the electric shocks (i.e. fewer electric shocks with better metacognitive performance). Our predictions were confirmed. Threat perceived as unrelated to metacognitive performance (Experiment 1) decreased metacognitive performance. Threat believed to be related to metacognitive performance resulted in improved performance (Experiment 2). These findings confirm that induced anxiety exerts a strong influence on metacognitive awareness while sparing first-order task performance. We furthermore demonstrate that the perceived relatedness of the anxiety determines whether metacognitive performance decreases or improves performance under threat.