It’s not what you expect: feedback negativity is independent of reward expectation and affective responsivity in a non-probabilistic task

Abstract

ERP studies commonly utilize gambling-based reinforcement tasks to elicit feedback negativity (FN) responses. This study used a pattern learning task in order to limit gambling-related fallacious reasoning and possible affective responses to gambling, while investigating relationships between the FN components between high and low reward expectation conditions. Eighteen undergraduates completed measures of reinforcement sensitivity, trait and state affect, and psychophysiological recording. The pattern learning task elicited a FN component for both high and low win expectancy conditions, which was found to be independent of reward expectation and showed little relationship with task and personality variables. We also observed a P3 component, which showed sensitivity to outcome expectancy variation and relationships to measures of anxiety, appetitive motivation, and cortical asymmetry, although these varied by electrode location and expectancy condition. Findings suggest that the FN reflected a binary reward-related signal, with little relationship to reward expectation found in previous studies, in the absence of positive affective responses.

Electrocortical components of anticipation and consumption in a monetary incentive delay task

“…In these studies, the RewP (or FRN) has typically been investigated as an electrocortical manifestation of reward prediction errors, with more negative amplitudes being observed for loss feedback than for gain feedback, and the difference between feedback valences being enhanced by greater reward expectancies (Bellebaum & Daum, ; Bellebaum, Polezzi, et al, ), and when outcomes differ from expectations (Pfabigan et al, ). However, these effects have not always been consistently observed, with several studies finding that the FRN is sensitive to binary differences in outcome (e.g., reward vs. nonreward) and not expectancy and expectancy violation (Highsmith, Wuensch, Tran, Stephenson, & Everhart, ). The findings of the present study are in broad agreement with research that interprets the RewP as a binary index of desired versus undesired outcomes, rather than a measure of expectancy violation and prediction error per se.…”