How attention and metacognitive biases influence perceptual decision-making in peripheral vision
Abstract
Project Summary: Peripheral vision plays a critical role in making accurate perceptual decisions: from driving on a crowded roadway,, to walking along a busy sidewalk, to identifying targets in settings with many distractors, visual information outside the center o,f our gaze contributes fast and useful data to guide decisions about where to look and where to move. While peripheral visions cont,ributions to perception and action are well-established, it is marked by paradoxical findings, as attempts to characterize how much,of the periphery we actually perceive have yielded conflicting evidence. Some findings indicate we perceive a surprising amount of d,etail, but limited working memory capacity constrains the number of items we can report. On the other hand, studies investigating in,attentional blindness and change blindness provide evidence that perception of unattended items in the periphery is limited. Thus, t,he rich vs. sparse debate surrounding peripheral vision pervades the scientific literature. Attempts to reconcile these discrepant, findings abound, but controversy remains.I have recently proposed a framework which may explain these results, called subjective i,nflation. Subjective inflation predicts that perceptual and metacognitive biases pervade decision-making in the visual periphery, a,nd cause observers to think they see more of the periphery than they really do. Specifically, peripheral representations suffer from, increased encoding noise in unattended areas, which causes (1) the tendency to think items are present, even when they are not (i.e,., false alarms), and (2) overconfidence in perceptual judgments. Evidence supporting this hypothesis is sound, but sparse. Therefor,e, the overall research goal of this proposal is to systematically evaluate how the tendency for high overconfidence/false alarms ex,tends across the periphery, to reveal how attention influences these biases, and to illuminate how visual context influences their p,revalence. Combining visual psychophysics, virtual reality, and real-time EEG to decode the locus of visual attention, the research,program outlined in this proposal will fill important gaps in our fundamental understanding of perceptual decision-making in the vis,ual periphery. Research Aim 1 in the proposed work uses virtual reality to characterize the presence of false alarms and high confid,ence across the periphery, and develop training protocols to reduce these biases. If experiments under Research Aim 1 are successful,, then we will have developed protocols to identify and reduce perceptual decision-making biases in sailors. Research Aim 2 in the p,roposed work incorporates real-time decoding of the focus of spatial attention using EEG, to determine how attention influences high, false alarms and overconfidence. If experiments under Research Aim 2 are successful, then we will have increased capability to unde,rstand how attentional allocation influences perceptual decision-making in target-based tasks, and we will be on the leading edge of, integrating these real-time advances in augmented reality systems, to improve human-computer interactions in visual decision-making,. Research Aim 3 investigates how different visual contexts influence these biases. If experiments under Research Aim 3 are successf,ul, then we will have a better understanding of how context influences decisional and metacognitive biases in sailors, to highlight,situations where these deficits occur most often. Together, these experiments will contribute basic scientific insights about the v,isual periphery and inform applications to improve pe
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 13, 2022
- Source ID
- N000142212534
Entities
People
- Brian Odegaard
Organizations
- Office of Naval Research
- United States Navy
- University of Florida