Metacognitive Confidence Increases with, but Does Not Determine, Visual Perceptual Learning

Abstract

While perceptual learning increases objective sensitivity, the effects on the constant interaction of the process of perception and its metacognitive evaluation have been rarely investigated. Visual perception has been described as a process of probabilistic inference featuring metacognitive evaluations of choice certainty. For visual motion perception in healthy, naive human subjects here we show that perceptual sensitivity and confidence in it increased with training. The metacognitive sensitivity–estimated from certainty ratings by a bias-free signal detection theoretic approach–in contrast, did not. Concomitant 3Hz transcranial alternating current stimulation (tACS) was applied in compliance with previous findings on effective high-low cross-frequency coupling subserving signal detection. While perceptual accuracy and confidence in it improved with training, there were no statistically significant tACS effects. Neither metacognitive sensitivity in distinguishing between their own correct and incorrect stimulus classifications, nor decision confidence itself determined the subjects’ visual perceptual learning. Improvements of objective performance and the metacognitive confidence in it were rather determined by the perceptual sensitivity at the outset of the experiment. Post-decision certainty in visual perceptual learning was neither independent of objective performance, nor requisite for changes in sensitivity, but rather covaried with objective performance. The exact functional role of metacognitive confidence in human visual perception has yet to be determined.

Conscious perception of flickering stimuli in binocular rivalry and continuous flash suppression is not affected by tACS-induced SSR modulation

“…What we observed were time-on-task effects independent of tACS. In the first CFS experiment, RT and accuracy improved over time, as could be expected (Zizlsperger et al, 2016). In the second CFS experiment, contrary to our expectation, metacognitive sensitivity decreased with time, an effect driven by a decrease in confidence.…”

Examining the robustness of the relationship between metacognitive efficiency and metacognitive bias

“…Given our current findings, it is possible that at least a portion of the metacognitive increase in that study was due to the increase in confidence, and indeed a follow-up paper that replicated Carpenter et al's design but used feedback that did not penalize low confidence did not find a change in metacognitive efficiency (De Gardelle et al, 2020). A number of other studies have examined how metacognition changes over the course of training (Bang et al, 2019;Guggenmos et al, 2016;Haddara & Rahnev, 2020;Schwiedrzik et al, 2011;Zizlsperger et al, 2016), or how it correlates with factors such as brain volume (Allen et al, 2017;Fleming et al, 2010;McCurdy et al, 2013;Rahnev et al, 2015) or age (Palmer et al, 2014), but typically did not try to control for potential effects of overall confidence (that is, metacognitive bias). To avoid confounding metacognitive bias and metacognitive efficiency, future studies should minimize or control for changes in one variable when examining the other.…”

Sensory noise increases metacognitive efficiency.

“…Rather, perceptual learning was used as a tool that allowed us to decrease sensory noise in our model. Several previous studies have combined confidence ratings and perceptual learning (Guggenmos, Wilbertz, Hebart, & Sterzer, 2016;Schwiedrzik, Singer, & Melloni, 2011;Solovey et al, 2016;Zizlsperger, Kümmel, & Haarmeier, 2016) but while they found important effects of learning on the overall confidence level, none investigated how training affects metacognitive efficiency.…”