Complexity and Dynamic Learning under Uncertainty: Experimental Studies

This project investigates how individuals learn and make decisions in environments characterized by uncertainty in hazard rates, with a particular focus on the role of structural complexity, operationalized through the number of informational inputs, asymmetry across signals, memory limitations, and framing of probabilistic signals, in shaping behavior. The project uses experimental methods to understand how people make these decisions when people do not know the chance of success and then tests economic theory using data from decisions made in these controlled environments. The project also evaluates whether and how interaction with an AI system can assist people in making decisions under conditions of uncertainty. The project develops a formal mathematical framework in which human agents learn about an unobserved hazard rate. This model is then tested in laboratory experiments. The experiments vary the structure of the information stream while keeping the actual information constant. The results identify when complex information environments lead to timely decisions. Experiments give participants access to an AI-based decision support tool to test how the tool interacts with human cognition. Overall, the project advances scientific knowledge of sequential learning and strategic decision-making. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2548818 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Byung-Cheol Kim | Institution: University of Alabama Tuscaloosa, TUSCALOOSA, AL | Award Amount: $520,547 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2548818 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2548818.html