Item

Abstract

Misinformation presents a challenge to societies worldwide, yet the cognitive computations underlying its processing remain scarcely understood. Here, we present a behavioural task and accompanying Bayesian models that allow us to study key aspects of the phenomenon and frame it as a learning problem about the trustworthiness of information providers. Specifically, we formulate a dual learning process where agents simultaneously learn about topics covered as well as the qualities of the news provider. In our task, participants are confronted with several different types of (mis-)information, ranging from a lying source to a source with biased reporting. Computational modelling of participants choices reveals both failures and successes of this learning process and sheds new light on how people come to trust, or distrust, a wide variety of information sources.