Electrocortical correlates of value-based decisions

Abstract

Choice behaviour is characterised by a calculation of subjective value of all choice options followed by the selection of the most subjectively valuable option. Neuroeconomics was developed to unify the fields of economics, psychology and neuroscience to describe neurobiological processes underpinning observed preferences. However, many aspects of the neural computations required to compute subjective value are yet to be illuminated. The current thesis aimed to describe the spatiotemporal dynamics of subjective valuation within the context of an auction paradigm. Brain processes related to subjective valuation were investigated using the incentive compatible Becker-DeGroot-Marschak auction task, revealing willingness-to-pay values for a variety of stimuli. Brain responses in response to stimulus onset and eye-fixations were extracted to highlight the electrophysiological response during value-based decisions, revealing the temporal evolution of brain processes on a scale of milliseconds. Analyses were accompanied by source dipole localisation methods to identify possible neural generators of observed responses. Data from four experimental chapters implicated the automatic encoding of various magnitudes of subjective willingness-to-pay within multiple cortical activation patterns during value-relevant and value-irrelevant contexts. Low value items were encoded most prominently by a brain activation component represented over the right frontal scalp region, possibly originating from the insula. In contrast, a cortical activation pattern with a spatial maximum over the left hemisphere was strongest for high value items. Brain components specific to value were present immediately following stimulus presentation and persisted throughout an extended time interval during free viewing. Further to this, value encoding brain responses were observed for both products and product bundles. The current thesis demonstrated the explicit representation of willingness-to-pay within neural activity recorded by means of EEG. Moreover, low and high subjective value was encoded by separate and lateralised brain responses in a coarse manner, in contrast to the linear encoding of subjective value within a single brain response. The early representation of value within EEG signals, prior to conscious deliberation, reiterates the automaticity of the initial valuation process.