Temporal Dynamics of Basic Decision-Making
Research in the Decision Neuroscience Lab uses high-density electroencephalography (EEG) to and advanced localization techniques to map how neural value signals emerge over the time course of choice. We have identified neural correlates of subjective value from approximately 450 to 650 ms after stimulus onset at central and frontotemporal sensors. This activity is localized to regions including the ventromedial prefrontal cortex (VMPFC), a region previously implicated in valuation and reward. Ongoing research in the lab examines how neural value signals are affected by other factors such as self-control*, physical effort*, and deciding for oneself* versus others. *See next slides
Regulation & Self-Control
Optimal decision making often requires that we exercise self-control in order to forego tempting but inferior immediate options for greater long-term rewards. One focus of our research is understanding how brain regions associated with executive control influence value signals in the VMPFC. Previous research on self-control suggests that the dorsolateral prefrontal cortex (DLPFC) plays an important role in this process. Examining the time course of self-regulation with ERP, we have found that DLPFC contributes to successful self-regulation in two different ways: 1) by reducing early perceptual processing of food stimuli, and 2) by modulating later value signals in the VMPFC (Harris et al., 2013).
VALUE & SENSORIMOTOR SYSTEMS
We have a lifetime of experience physically interacting with objects that we value, and acquiring them often requires effort. How does the brain calculate the trade-off between effort cost and reward? Many cognitive models of choice assume that physical effort factors into decisions relatively late, whereas research on embodied cognition suggests that sensorimotor systems may be engaged from early in the decision process. Mapping brain activity during decisions associated with physical effort, we found that rapid sensorimotor activation is integrated with cognitive representations of value in the time leading up to motor response (Harris & Lim, 2016). Additionally, specific directional actions (e.g., reaching towards an object) are associated with conceptual representations of item value (Green et al., 2017), suggesting that action and value are inextricably linked.
Social & Financial Decision-Making
Decision making is often more complicated than simply picking what we like for ourselves, whether we’re making lunch for our children or making medical or legal decisions as proxies for others. How do we choose for others, particularly when their preferences differ from our own? Combining ERP with computational modeling, we found that integration of others’ preferences occurs over the same timescale as for oneself, while differentially tracking recipient-relevant attributes (Harris, Clithero & Hutcherson, in preparation). Financial trading presents another complex decision scenario where investors must take into account both the current valuation of a stock and whether it is gaining or losing value. In a joint project with Drs. Cary Frydman and Tom Chang at the USC Marshall School of Business, we have collected brain activity while people made investments in an experimental stock market. Preliminary results suggest that the VMPFC integrates the value of capital gains on a similar timescale as other goods like food, and this activity is correlated with the propensity to sell winning stocks too soon (the “disposition effect”).