The spatio-temporal dynamics of value-based action-selection
We make countless choices every day, and we expect them to lead to favorable outcomes. Past research on such value-based decision-making focused largely on valuation and reinforcement learning processes, while action-selection received surprisingly little attention. This is despite the fact that without selection, there will often be no reward at all. We know that selection requires an intricate interplay of cortical and subcortical brain regions. Yet, models of value-based decision-making mostly assume that selection simply occurs optimally. In this project, I argue that this view is much too simplistic, and propose a research agenda to systematically investigate the neural processes underpinning value-based action-selection. I will focus on the common assumption that valuation and action-selection are two separate processes implemented in separate brain regions. This view is counter to past research and current theories of action-selection, and we will test whether and how valuation affects selection processes in the brain. We will use a combination of state-of-the-art computational models, multivariate pattern analysis techniques, fMRI, and EEG, to comprehensively describe the spatial and temporal dynamics of value-based action-selection. Through its sure footing in current theories, and innovative use of novel analysis methods, this project will significantly impact our understanding of decision-making, and adaptive behavior more generally.