The Dynamics of Unique Decisions

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

The fields of decision neuroscience and neuroeconomics have seen a dramatic increase in the use of computational modeling to advance understanding of people’s predispositions and choice biases. The models use choice-process data consisting of responses such as eye movements during the evaluation of the options, hand movements during the execution of the choice, and the time that it takes to decide. However, there remains substantial skepticism about the usefulness of these methods for understanding decisions such as buying cars or choosing retirement funds. A major obstacle has been that complex or high-stakes decisions are very challenging to study with these methods. This project develops novel ways to circumvent such problems and collects choice-process data online and applies computational modeling to one-shot decisions. This enables modeling of the choice process in online experiments while leveraging large online samples to eliminate the need for designs with repeated decisions. This research focuses on the extent to which behavior is driven by biases that people bring to their decisions. Additionally, the project investigates how people interpret and manipulate decision times in strategic interactions. Overall, this work extends the scope of computational modeling of choice process data to complex, high-stakes decisions, with decision times much longer than typically studied. The education and outreach parts of the project entail many activities such as research internships, summer institutes, and conference workshops designed to introduce students of all ages, and those with disabilities, to the exciting decision science research and motivate their pursuit of careers in STEM. This project entails research, education, and outreach activities focused on dynamic modeling of the choice process in complex, high-stakes human decisions. The research features an innovative approach to study one-shot (“unique”) decisions, by combining novel computational modeling approaches with choice-process data (eye tracking, mouse tracking, and response times) collected from online experiments, existing market data, and field experiments. The project develops and tests techniques for single-trial computational modeling and online eye-tracking; develops a mapping between features of computer mouse trajectories and choice-process components; and studies choice processes using data from online markets. The project also implements a multi-faceted broader impacts plan aimed at increasing interest, participation, literacy, and retention of students in the decision sciences and STEM more broadly. The project offers summer research internships to undergraduate students with disabilities; hands-on decision-science activities for high-school students in the Ohio Supercomputer Center’s (OSC) summer institute; an annual visit to the Ohio State experimental economics lab for inner-city middle-school students; an annual workshop on “choice-process data” for experimental economists; and development of a toolbox for online webcam-based eye-tracking technology. 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.