bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 12, 2024

Abstract Humans are prone to decision biases, which make behavior seemingly irrational. An important cause for biases is that the context in decisions made can later influence choices humans prefer new situations. Current computational models often require extensive environmental knowledge explain these biases. Here, we tested hypothesis mainly driven by a tendency repeat context-specific actions. We implemented series of nine value-based experiments on n = 351 male and female participants reanalyzed six previously published datasets (n=350 participants). found higher within-context repetition an option was associated with biased including subjective valuation lower uncertainty repeated Next, used model based two basic principles, learning reward it all datasets. Our results show combination principles sufficient choices. demonstrate via hierarchical Bayesian comparison our outperforms alternative models. These provide novel insights into human choice offer solution difficulties current explanations Significance Statement Human decision-making deviates from rational theory. Understanding how emerge essential interpreting real-world behavior. study shows repeating shapes making adapt simple repetition. By testing across 15 datasets, previous models, also offering preferences affect uncertainty. In sum, findings deeper understanding context-dependent specific persist.

Language: Английский