Abstract

It is generally believed that humans can detect errors in their estimates and maintain a sense of confidence that corresponds to objective performance. Meanwhile, human perceptual estimations tend to commit two types of errors: bias and variability. It remains unclear whether humans can monitor these two distinct kinds of error in their own estimates and adjust levels of confidence accordingly. Using a well-established empirical regularity called the central tendency effect, we examined how behavioral bias and variability are reflected in the self-evaluation of perceptual estimations. Subjects estimated the location of a hidden target from a briefly presented dot-cloud centered on the target location and reported levels of confidence in their estimations. As expected, subjects' perceptual reports were not only variable across trials but also biased toward the mean of the distribution of target locations. Crucially, subjects' confidence reports remained constant regardless of the target location, which is in stark contrast to the fact that the perceptual performance was worse when the target location was farther from the mean of the target location distribution. Our findings indicate that the metacognitive evaluations of perceptual estimations reflect the behavioral variability, but not biases.