Go to ACL ARR 2025 February homepageHow Does Cognitive Bias Affect Large Language Models? A Case Study on the Anchoring Effect in Price Negotiation Simulations
Abstract: Cognitive biases, well studied in humans, can also be observed in LLMs, affecting their reliability in real-world applications. This paper investigates the anchoring effect in LLM-driven price negotiations. To this end, we instructed seller LLM agents to apply the anchoring effect and evaluated negotiations using not only an objective metric but also a subjective metric. Experimental results show that LLMs are influenced by the anchoring effect like humans. Additionally, we investigated the relationship between the anchoring effect and factors such as deliberation and personality. It was shown that deliberation mitigates the anchoring effect, suggesting that more deliberative models are less prone to the effect. However, we found no significant correlation between personality traits and susceptibility to the anchoring effect. These findings contribute to a deeper understanding of cognitive biases in LLMs and provide insights into their implications for real-world applications.
Paper Type: Long
Research Area: Linguistic theories, Cognitive Modeling and Psycholinguistics
Research Area Keywords: cognitive modeling, computational psycholinguistics
Contribution Types: Publicly available software and/or pre-trained models, Data analysis
Languages Studied: English
Submission Number: 1012
Loading