Abstract
Introduction: Visuomotor strategies come close to maximising expected gain but economic preference tasks indicate that people often cognitively misrepresent the probability of outcomes (Kahneman & Tversky, 1972; Trommershäuser et al., 2008). One explanation for the cognitive biases is that individuals evaluate outcomes relative to prior reward probability contexts rather than the current objective value of the feedback. The present study examined the neural correlates of relative changes in reward probability. Methods: We measured the visual feedback-related negativity (FRN) in a gambling task. The FRN is an electrophysiological correlate of attention-related outcome processing, which differentiates between rewards and punishments (Gehring & Willoughby, 2002). On each trial, participants decided between a risky and a safe option, presented as different coloured shapes. Each gamble was accompanied with visual feedback. The reward probability (RP) associated with gambles was manipulated so there were four different blocks: (i) ‘Win Block’ (WB) with a positive context of 80% RP; (ii) ‘Loss Block’ (LB) with a negative context of 20% RP; (iii) ‘Post-Win’ (PW) block of 50% RP occurring immediately after a WB; (iv) ‘Post-Loss’ block of 50% RP occurring immediately after a LB. Results: 50% RP environments yielded different FRN patterns according to whether they reflected a relative improvement or decrement from the previous condition. This is consistent with the idea that outcome-monitoring processes are influenced by relative states. The pattern of FRN activity in the PL and PW blocks mirrored activity in the WB and LB respectively; suggesting the FRN is driven by subjective perceptions of outcome. Conclusions: The data indicate that relative changes in reward probability from previously learned contexts can bias our evaluation of the valence of current outcomes. These neurophysiological data shed some light on the factors that produce differences between visuomotor and cognitive strategies in decision-making.
Meeting abstract presented at VSS 2013