arXiv q-fin· Pedro Gonzalez-Fernandez·· 13 hr agoAI Score40
Identifying Belief Biases
Belief Bias Identification
AI summary
This study proposes a unified theoretical model that can identify and test a full set of probability updating biases within a single framework, focusing on belief distribution updates rather than point beliefs. Estimates from laboratory experiments show that the hot-hand fallacy is the most common bias and a key driver of biased reasoning. Most biases cancel out at the aggregate level, but base-rate neglect persists.
Source:arXiv q-fin · arxiv.org