A Gambler's Fallacy for Probability Judgments when Event Sequences Are Truly Random: A Reanalysis of Xiang, Dorst, and Gershman's (2025) Data.

This article focuses on recent research challenging and reanalyzing theories related to the gambler's fallacy in psychological sciences. Xiang, Dorst, and Gershman (2025) argued that the gambler's fallacy appears in binary predictions but not in probability judgments when event sequences are truly random, suggesting it does not stem from probabilistic reasoning. However, a subsequent reanalysis of their data found strong evidence of the gambler's fallacy in probability judgments and a connection between probability judgments and binary predictions. Improved models, including an enhanced representativeness model and Rabin and Vayanos's (2010) model, were shown to account for both response types, indicating that probabilistic theories remain viable. The article also proposes stratified random sampling as a more efficient method for presenting event sequences in such studies. [Extracted from the article]