Neurocomputational Mechanisms of Affiliation and Personality Study (NeuroMAP):
Some people struggle with emotional instability, stormy relationships, and impulsive or self-harming behaviors. Our NeuroMAP study seeks to understand how these challenges relate to decision processes that can be measured by functional neuroimaging and computational reinforcement learning models. This approach builds on neuroscience research that has identified a key distinction between Pavlovian and goal-directed neurocomputational systems that are implemented in specific brain circuits.
Our study will examine whether circuits involved in goal-directed learning are vulnerable to disruptions by social and emotional cues that exert Pavlovian influences on decision-making. This work can help us understand some kinds of mental illness such as borderline personality and social anxiety disorder.
The study will also consider how the maturation of control-related brain networks influences Pavlovian and goal-directed systems. The NeuroMAP study is funded by an award to Dr. Hallquist from the National Institute of Mental Health Biobehavioral Research Awards for Innovative New Scientists (NIMH BRAINS) program.
When something unexpectedly good happens, our mood often increases. On the other hand, when negative things happen, we often feel sad or disheartened. Although psychology has long noted the reciprocal relationships between mood and reward learning, the field lacks formal mathematical models that can potentially explain individual differences in emotional dynamics. This is an important challenge in part because unstable and sometimes extreme emotions are a common feature of many forms of mental illness.
Our Momentum study aims to describe the relationships among reward responsiveness, reinforcement learning, and subjective valuation using a new computational reinforcement learning model.
Using assessments of physiology, brain activity, reward responsiveness, and subjective mood in daily life, we hope to provide new insights into the neurobiological and psychosocial underpinnings of emotional instability.
This work is funded by the National Institute of Mental Health as part of its Research Domain Criteria (RDoC) initiative to understand broader biological and cognitive processes associated with mental health and illness.
The B-Social study examines social decision-making in individuals with borderline personality disorder (BPD). Tumultuous interpersonal relationships and difficulties with social decisions are hallmarks of BPD. We think of social interactions in the framework of reinforcement learning, where social information is tracked alongside other sources of rewards and punishments to inform an integrated representation of subjective value. For this project, we are especially interested in the interplay of Pavlovian and instrumental representations during social decision-making. Data collection for the B-Social study occurs at the University of Pittsburgh as part of a longitudinal study of BPD. This research occurs in close collaboration with the Decision Neuroscience and Psychopathology Lab at Pitt (Director: Alexandre Dombrovski).
Emotions and Decision Making Study (EDMS):
Emotions have a potent effect on the types of decisions we make (e.g., we may hug a friend if we’re feeling caring). Yet even as this phenomenon is widely discussed in popular culture, the exact ways in which emotions affect decision making are not well understood.
EDMS is interested in understanding the precise ways in which emotions alter different types of decisions. We’ve developed a novel task that allows us to observe some of these types of decisions so that we can begin to understand this phenomenon.
In this study, participants are asked to complete a computer game in which they’re tasked with navigating a space (e.g., a fictional town) and engaging in actions (e.g., buying a book at the bookstore) in order to achieve a goal (e.g., delivering a book to a friend). Some participants receive different task manipulations, which allow us to understand the influence of emotions on these decisions. We use computational modeling of participants’ task behavior to quantify how these task manipulations alter different decision processes.