Neuropsychologia 46 (2008) 521–539 Better or worse than expected? Aging, learning, and the ERN Ben Eppinger ∗ , Jutta Kray, Barbara Mock, Axel Mecklinger Department of Psychology, Saarland University, Campus Building A1 3, D-66041 Saarbr¨ ucken, Germany Received 20 February 2007; received in revised form 23 August 2007; accepted 4 September 2007 Available online 7 September 2007 Abstract This study examined age differences in error processing and reinforcement learning. We were interested in whether the electrophysiological correlates of error processing, the error-related negativity (ERN) and the feedback-related negativity (FRN), reflect learning-related changes in younger and older adults. To do so, we applied a probabilistic learning task in which we manipulated the validity of feedback. The results of our study showed that learning-related changes were much more pronounced (a) in a response-locked positivity for correct trials compared to the ERN and (b) in a feedback-locked positivity for positive feedback compared to the FRN. These findings provide an important extension to recent theoretical accounts [Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human error processing: Reinforcement learning, dopamine, and the error-related negativity. Psychological Review, 109, 679–709; Nieuwenhuis, S., Ridderinkhof, K. R., Talsma, D., Coles, M. G. H., Holroyd, C. B., Kok, A., et al. (2002). A computational account of altered error processing in older age: Dopamine and the error-related negativity. Cognitive, Affective and Behavioral Neuroscience, 2, 19–36] since they suggest that positive learning signals on correct trials contribute to the reward-related variance in the response- and feedback-locked ERPs. This effect has been overlooked in previous studies that have focused on the role of errors and negative feedback for learning. Importantly, we did not find evidence for an age-related reduction of the ERN, when controlling for performance differences between age groups, which questions the view that older adults are generally impaired in error processing. Finally, we observed a substantial reduction of the FRN in the elderly, which indicates that older adults are less affected by negative feedback and rely more on positive feedback during learning. This finding points to an age-related asymmetry in the processing of feedback valence. © 2007 Elsevier Ltd. All rights reserved. Keywords: Age differences; ACC; Dopamine; ERP; Error processing; Reinforcement learning 1. Introduction The aim of this study is to examine age-related modulations in reinforcement learning and their potential impact on the ability to adaptively acquire and maintain new behavior. We focus on the role that the processing of error information plays for rein- forcement learning by investigating age differences in the behav- ioral and electrophysiological correlates of error processing. The basic principle of reinforcement learning has already been formulated in the early 1920s by Thorndike. It suggests that actions that are followed by feelings of satisfaction are more likely to be generated again in the future, whereas actions that are followed by negative outcomes are less likely to reoccur (Thorndike, 1911). In recent years several researchers became interested in the neurophysiological basis of reinforcement ∗ Corresponding author. Tel.: +49 681 3023698; fax: +49 681 3023871. E-mail address: eppinger@mx.uni-saarland.de (B. Eppinger). learning and especially in the role of dopamine for learn- ing. For instance, electrophysiological studies in primates have shown that learning based on rewarding outcomes depends on the activity of dopaminergic neurons in the ventro-anterior midbrain (substantia nigra and ventral tegmental area) (for reviews, see Montague, Hyman, & Cohen, 2004; Schultz, 2002). Schultz, Dayan, and Montague (1997) integrated these electro- physiological findings in primates with learning models from artificial intelligence. They showed that changes in the activity of dopaminergic neurons during learning could be formalized using a temporal difference learning model (see Barto & Sutton, 1997). Dopaminergic neurons from the ventral tegmental area seem to signal the extent to which a rewarding outcome deviates from a prediction during learning. That is, they code prediction errors that reflect changes in the value of ongoing events, when events are suddenly better or worse than expected (Schultz et al., 1997). According to this model, learning is induced when a reinforcer occurs that is better than predicted (positive prediction error). In contrast, a reinforcer that is worse than predicted or 0028-3932/$ – see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2007.09.001