Individual Differences In Value-Based Decision-Making: Learning And Time Preference

Human decisions are strongly influenced by past experience or by the subjective values attributed to available choice options. Although decision processes show some common trends across individuals, they also vary considerably between individuals. The research presented in this dissertation focuses on two domains of decision-making, related to learning and time preference, and examines factors that explain decision-making differences between individuals. First, we focus on a form of reinforcement learning in a dynamic environment. Across three experiments, we investigated whether individual differences in learning were associated with differences in cognitive abilities, personality, and age. Participants made sequential predictions about an on-screen location in a video game. Consistent with previous work, participants showed high variability in their ability to implement normative strategies related to surprise and uncertainty. We found that higher cognitive ability, but not personality, was associated with stronger reliance on the normative factors that should govern learning. Furthermore, learning in older adults (age 60+) was less influenced by uncertainty, but also less influenced by reward, a non-normative factor that has substantial effects on learning across the lifespan. Second, we focus on delay discounting, the tendency to prefer smaller rewards delivered soon over larger rewards delivered after a delay. Delay discounting has been used as a behavioral measure of impulsivity and is associated with many undesirable real-life outcomes. Specifically, we examined how neuroanatomy is associated with individual differences in delay discounting in a large adolescent sample. Using a novel multivariate method, we identified networks where cortical thickness varied consistently across individuals and brain regions. Cortical thickness in several of these networks, including regions such as ventromedial prefrontal cortex, orbitofrontal cortex, and temporal pole, was negatively associated with delay discounting. Furthermore, this brain data predicted differences beyond those typically accounted for by other cognitive variables related to delay discounting. These results suggest that cortical thickness may be a useful brain phenotype of delay discounting and carry unique information about impulsivity. Collectively, this research furthers our understanding of how cognitive abilities, brain structure and healthy aging relate to individual differences in value-based decision-making