Beyond Traditional Dual-Process Theories of Social Cognition: Using Reinforcement Learning to Understand Attitude Change, Moral Reasoning, and Impression Formation

We highlight how reinforcement learning models, rooted in animal learning and formalized in computer science, offer a computationally tractable alternative to traditional dual-process theories in understanding value-based learning and choice and help move toward theoretical unification across different areas of social cognition, including attitude change, moral reasoning, and impression formation.