UC Berkeley

The ability to initiate goal-directed behaviors related to every day outcomes is a crucial component of motivation. Across clinical and subclinical domains, variability in motivated decision-making accounts for failures to both begin and complete actions towards improved health. When faced with decisions that involve expending our own energetic resources in order to attain rewards, whether for ourselves or others, individuals vary broadly in the behaviors they take, or fail to take, to reach said goals. In order to classify, and potentially provide a mechanistic account of, differences in these cost-benefit processes, an objective investigation into the computations underlying motivated behavior is tantamount. This is particularly relevant to the clinical domain, wherein the broad goal of computational psychiatry is a methodologically quantitative approach to taxonomic classification. The following essays offer an overview of three studies with the aim of providing formal accounts of integrating information from individual differences (chapters 1 and 2) and stereotype content (chapter 3) into mathematical models of motivated behavior. I will begin by discussing a computational approach towards predictive modeling of effort-based decision-making utilizing dimensions of clinical and subclinical approach and avoidant traits. In the first chapter, I will focus on modeling behavior in a community-recruited sample of adults by integrating a number of demographic and individual difference measures into behavioral economic models of effort discounting. In the second, I will focus my analyses on a separate but demographically similar cohort of adults who completed the same experimental paradigm during functional neuroimaging. Here, I will investigate changes in neural activity (inferred by blood-oxygen level dependent (BOLD) signal) responding to different levels of effort-reward tradeoffs. I will utilize the same behavioral modeling processes as the first chapter in order to extract individual difference measures by estimating free parameters specific to individual subjects, and relating them to changes in the neural responses to cost-benefit evaluations. Finally, I will discuss the manner in which evaluations of effort expenditure for other-rewarding outcomes are distinct from those with self-rewarding outcomes, and how social information – or stereotypes – about the recipient of said rewards robustly contributes to effort-related choice processes