Metacognitive ability limits visual attentional control

Undergraduate Just-In-Time Abstract

Poster Presentation: Tuesday, May 20, 2025, 2:45 – 6:45 pm, Banyan Breezeway
Session: Undergraduate Just-In-Time 2

Sydney Trowbridge1, Owen Vegas2, Michael Grubb2, Silvia Lopez-Guzman1; 1Unit on Computational Decision Neuroscience, National Institute of Mental Health, National Institutes of Health, Bethesda, MD., 2Psychology Department, Trinity College, Hartford, CT.

Attentional Blink (AB) is the decreased ability to detect a second target when it follows an initial target by ~200ms and has been widely studied in perceptual neuroscience. AB appears to improve after engaging in metacognition through a meditation intervention. We hypothesized that these training effects could be limited by differences in individuals’ ability to introspect. We addressed this question by measuring AB before and after completing a perceptual decision-making task where participants had to introspect and report their decision confidence. 72 participants completed the experiment online. In the AB task they had to report on the presence of a second target (T2) after identifying an initial target (T1), in a rapid stream of visual stimuli. In the metacognition task, participants were presented with four visual gratings, with three oriented vertically at 0° and one tilted to the left or right by a range of 0.5°-30°. Participants indicated with a keystroke whether the orientation of the tilted Gabor was clockwise or counterclockwise, then reported their confidence that they were correct on a 1-4 scale. AB data was fit to a signal detection theory (SDT) based linear model through hierarchical Bayesian estimation, to derive individual-level decision criterion and decision sensitivity and their change pre and post confidence task. Confidence data was fit to the CASANDRE model (also rooted in SDT) to derive individual confidence sensitivity. SDT modeling revealed individuals with greater metacognitive ability exhibited higher decision sensitivity in the baseline AB task. However, we found no significant relation between hit rate in the AB task and confidence sensitivity either at baseline or pre vs. post, displaying the confounds associated with standard hit-rate analyses commonly used in AB literature. Given that better metacognition is related to enhanced attentional control, future studies could address whether introspection and visual attention share common mechanisms.

Acknowledgements: This research was supported (in part) by the Intramural Research Program of the NIMH - ZIAMH002988