AUTHOR=Al Dowaji Rawa , Xu Ji , Jin Yimeng , Porte Antoine , Lauwereyns Johan TITLE=The limits of metacognitive control during perceptual decision-making: opting out without improving accuracy JOURNAL=Frontiers in Psychology VOLUME=Volume 16 - 2025 YEAR=2025 URL=https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2025.1551665 DOI=10.3389/fpsyg.2025.1551665 ISSN=1664-1078 ABSTRACT=The purpose of this study was to examine how the inclusion of an opt-out option affects the metacognitive control of perceptual decision-making under challenging conditions. In Experiments 1 and 2, participants were required to compare the flicker frequency of simultaneously presented stimuli. In Experiment 3, participants had to identify the dominant color in a patch of red and green dots. We hypothesized that, with an option to skip, participants would strategically opt out of trials in which they were uncertain, thus reducing their error rates and improving their overall performance. In Experiments 1 and 2, we compared conditions under time pressure with versus without a skip option. We also varied the risk, or penalty associated with error. By raising the risk, we found that participants tended more often to opt out of the decision. However, this escape behavior did not enable them to achieve better performance. The opt-out decision appeared to impose a cognitive burden, requiring additional effort without yielding a clear advantage. In Experiment 3, we varied the time pressure with a short versus long deadline. We also manipulated the task difficulty with color dot ratios that were easy or hard to discriminate. Participants tended to skip more often in hard trials than in easy trials, whereas the short versus long deadline did not affect the skip rate. Again, the increase in opting out did not lead to reduced error rates. Across the three experiments, we found that factors such as risk and task difficulty elicited escape behavior in perceptual decision-making, without improving accuracy. Thus, the participants demonstrated they could monitor their performance but were unable to achieve strategic metacognitive control with the opt-out option.