Cognitive science commonly assumes human learners combine prior knowledge with new information to form statistically optimal beliefs. Why then are false beliefs so resilient and widespread? We develop and test a new hypothesis stemming from computational models of human category learning. In four experiments we found that learners weight conflicting information sources pursuant to a nonlinear function dependent on the learners’ initial beliefs. Highly agreeable sources receive strong weight preventing belief change, but this deference drops sharply and then flattens with increasing disparity, allowing even very distal sources to effect belief change. To understand the implications of this bias for the maintenance of false beliefs in society, we simulated groups of agents learning from one another and from a truth-telling oracle. The learning bias often led agents to fractionate into mutually distrustful groups or to converge on a common false belief. Thus, opposing belief bubbles in contemporary society may reflect stable states in a dynamic system composed of individual learners communicating their beliefs to one another and updating those beliefs according to an evidence-weighting function that strongly favors close agreement but, in the absence of such agreement, never completely discounts extreme voices.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.