Goals as reinforcers

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

From collecting stamps to climbing mountains, many human activities are driven by goals rather than objective rewards. Oftentimes, people incur significant costs for the sake of achieving their goals, even when the direct, objective benefits of attaining them are limited – demonstrating people’s extraordinary ability to set and pursue their own challenges. This project investigates the cognitive mechanisms that allow individuals to leverage the brain’s motivation system for arbitrary goals. In particular, the project investigates how learning and decision-making differ when people attempt to obtain familiar, objective rewards, such as points, compared to novel, otherwise neutral goals. Discovering how differences between the two modalities arise should help clarify how people attribute value to otherwise neutral or even negative outcomes (such as a growling stomach) for the sake of reaching their objectives (for example, losing weight). By understanding the extent to which humans can use goals to direct their own learning and the factors that make it less efficient to do so, this research helps to develop novel techniques to support goal achievement in settings ranging from education and the workplace to public policy and personal development. Human learning through experience is typically understood as being driven by the expectation of outcomes that either are directly rewarding (such as water and food) or have been associated with primary rewards over time (such as money or numeric points). However, humans have a demonstrated ability to pursue even complex and effortful activities in hopes of attaining abstract, arbitrary, and often self-imposed goals. This project aims to uncover some of the cognitive machinery that enables human-level flexibility in goal-driven behavior. More specifically, this work focuses on understanding the potentially joint, yet distinct contributions of executive function and internal simulation processes on goal-congruent value attribution. To this end, the investigators use novel experimental designs to test people’s learning and decision-making under more or less abstract incentives. Paired with computational modeling, this approach enables the identification of key cognitive contributors to efficient goal achievement. The project also broadens participation in computational cognitive science, through the advanced technical training afforded by this project. The broader impacts of this proposal also include how the insights from this research may enable the development of tools supporting people in the achievement of important goals – from quotidian tasks to ambitious missions. 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.