More Like this

1. The effects of base rate neglect on sequential belief updating and real-world beliefs

   https://doi.org/10.1371/journal.pcbi.1010796  

   Ashinoff, Brandon K; Buck, Justin; Woodford, Michael; Horga, Guillermo (December 2022, PLOS Computational Biology)

   Gershman, Samuel J (Ed.)

   Base-rate neglect is a pervasive bias in judgment that is conceptualized as underweighting of prior information and can have serious consequences in real-world scenarios. This bias is thought to reflect variability in inferential processes but empirical support for a cohesive theory of base-rate neglect with sufficient explanatory power to account for longer-term and real-world beliefs is lacking. A Bayesian formalization of base-rate neglect in the context of sequential belief updating predicts that belief trajectories should exhibit dynamic patterns of dependence on the order in which evidence is presented and its consistency with prior beliefs. To test this, we developed a novel ‘urn-and-beads’ task that systematically manipulated the order of colored bead sequences and elicited beliefs via an incentive-compatible procedure. Our results in two independent online studies confirmed the predictions of the sequential base-rate neglect model: people exhibited beliefs that are more influenced by recent evidence and by evidence inconsistent with prior beliefs. We further found support for a noisy-sampling inference model whereby base-rate neglect results from rational discounting of noisy internal representations of prior beliefs. Finally, we found that model-derived indices of base-rate neglect—including noisier prior representation—correlated with propensity for unusual beliefs outside the laboratory. Our work supports the relevance of Bayesian accounts of sequential base-rate neglect to real-world beliefs and hints at strategies to minimize deleterious consequences of this pervasive bias. 

   more » « less
2. Crowdsourcing to analyze belief systems underlying social issues

   Priniski, J. A.; Holyoak, K. J. (January 2020, Proceedings of the Annual Conference of the Cognitive Science Society)

   People’s beliefs and attitudes about social and scientific issues, such as capital punishment and climate change, appear to form complex but generally coherent networks. Understanding the nature of these networks is a prerequisite for designing interventions for changing beliefs on the basis of rational arguments and evidence. It is therefore important to develop methods to represent and analyze the form and nature of belief networks, which may not be explicitly verbalizable. Adopting an emerging approach that utilizes crowdsourcing to develop educational interventions, we mined discussions from the Reddit forum Change My View to determine which beliefs and types of information underlie people’s attitudes about capital punishment. By combining computational analyses based on a topic model with more qualitative assessments of the extracted topics, we found that moral arguments are more prevalent than statistical or data-based arguments. The present study serves as a test case for the open sourced software crowdpy, a Python toolkit for running naturalistic studies on the web, which will enable other researchers to use crowdsourcing in their research. This approach sets the stage for research exploring potential interventions to change people’s beliefs. 

   more » « less
3. Strategic play by resource-bounded agents in security games

   X. Liu; J. Y. Halpern (May 2023, Proceedings of the 22nd International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023))

   Many studies have shown that humans are ``predictably irrational'': they do not act in a fully rational way, but their deviations from rational behavior are quite systematic. Our goal is to see the extent to which we can explain and justify these deviations as the outcome of rational but resource-bounded agents doing as well as they can, given their limitations. We focus on the well-studied ranger-poacher game, where rangers are trying to protect a number of sites from poaching. We capture the computational limitations by modeling the poacher and the ranger as probabilistic finite automata (PFAs). We show that, with sufficiently large memory, PFAs learn to play the Nash equilibrium (NE) strategies of the game and achieve the NE utility. However, if we restrict the memory, we get more ``human-like'' behaviors, such as probability matching (i.e., visiting sites in proportion to the probability of a rhino being there), and avoiding sites where there was a bad outcome (e.g., the poacher was caught by the ranger), that we also observed in experiments conducted on Amazon Mechanical Turk. Interestingly, we find that adding human-like behaviors such as probability matching and overweighting significant events (like getting caught) actually improves performance, showing that this seemingly irrational behavior can be quite rational. 

   more » « less
4. Extending rational models of communication from beliefs to actions

   Sumers, T.; Hawkins, R.D.; Ho, M.; Griffiths, T.L. (January 2021, Proceedings of the Annual Conference of the Cognitive Science Society)

   null (Ed.)

   Speakers communicate to influence their partner's beliefs and shape their actions. Belief- and action-based objectives have been explored independently in recent computational models, but it has been challenging to explicitly compare or integrate them. Indeed, we find that they are conflated in standard referential communication tasks. To distinguish these accounts, we introduce a new paradigm called signaling bandits, generalizing classic Lewis signaling games to a multi-armed bandit setting where all targets in the context have some relative value. We develop three speaker models: a belief-oriented speaker with a purely informative objective; an action-oriented speaker with an instrumental objective; and a combined speaker which integrates the two by inducing listener beliefs that generally lead to desirable actions. We then present a series of simulations demonstrating that grounding production choices in future listener actions results in relevance effects and flexible uses of nonliteral language. More broadly, our findings suggest that language games based on richer decision problems are a promising avenue for insight into rational communication. 

   more » « less
5. Networks of beliefs: An integrative theory of individual- and social-level belief dynamics.

   https://doi.org/10.1037/rev0000494  

   Dalege, Jonas; Galesic, Mirta; Olsson, Henrik (September 2024, Psychological Review)

   We present a theory of belief dynamics that explains the interplay between internal beliefs in people’s minds and beliefs of others in their external social environments. The networks of belief theory goes beyond existing theories of belief dynamics in three ways. First, it provides an explicit connection between belief networks in individual minds and belief dynamics on social networks. The connection, absent from most previous theories, is established through people’s social beliefs or perceived beliefs of others. Second, the theory recognizes that the correspondence between social beliefs and others’ actual beliefs can be imperfect, because social beliefs are affected by personal beliefs as well as by the actual beliefs of others. Past theories of belief dynamics on social networks do not distinguish between perceived and actual beliefs of others. Third, the theory explains diverse belief dynamics phenomena parsimoniously through the differences in attention and the resulting felt dissonances in personal, social, and external parts of belief networks. We implement our theoretical assumptions in a computational model within a statistical physics framework and derive model predictions. We find support for our theoretical assumptions and model predictions in two large survey studies (N1 = 973, N2 = 669). We then derive insights about diverse phenomena related to belief dynamics, including group consensus and polarization, group radicalization, minority influence, and different empirically observed belief distributions. We discuss how the theory goes beyond different existing models of belief dynamics and outline promising directions for future research. 

   more » « less