Value-based prioritization in human sensorimotor decision making
In challenging, dynamic environments, split-second sensorimotor decisions must be prioritized according to potential payoffs. An extensive literature has reported and modeled value-related biases in the timing and accuracy of choice behavior when subjects make deliberative perceptual judgments about weak stimuli. However, little is known about the mechanisms of adaptive value-biasing in situations where evidence is strong but the time to act is severely limited, such as in sports or when traveling in traffic. Moreover, no study has examined the neural processes involved in setting these biases in place in advance of the imperative sensory event. In my talk I will present the findings of two recent studies that address these questions using psychophysics, computational modeling and electrophysiology in humans performing a time-constrained, value-biased sensory discrimination task. In one study we identified a novel neural signature of preparatory value-based prioritization that scales with relative value and strongly predicts behavioral biases in the upcoming decision, and which bears interesting distinctions from hitherto known markers of attention, intention, salience, value or priority. In a second study, we examined mechanisms of value-based prioritization in the decision process itself. Behavioral model fits as well as neural motor preparation dynamics were consistent with value-biases being exerted at the sensory level, contrary to the vast majority of findings from paradigms emphasising perceptual uncertainty. The findings of these studies present a new model framework that furnishes concrete predictions for many other sensory stimuli and task scenarios, and provide a novel paradigm with considerable clinical potential, allowing principled assessment of how reward information is managed in the context of externally-instructed behaviors.