Response-related negativities following correct and incorrect responses: Evidence from a temporospatial principal component analysis TANJA ENDRASS, JULIA KLAWOHN, ROSA GRUETZMANN, MORITZ ISCHEBECK, and NORBERT KATHMANN Department of Psychology, Humboldt-Universität zu Berlin, Berlin, Germany Abstract Two ERP components have been observed following correct and incorrect responses, the error-related negativity (ERN/Ne) and the correct-related negativity (CRN). The function of these components is still under debate. We used a visual size discrimination task at three difficulty levels and utilized a temporospatial principal component analysis (PCA) to examine whether ERN/Ne and CRN could be explained by one or more factors. While ERN/Ne decreased with higher task difficulty, amplitudes increased for correct responses at parietal electrodes. PCA revealed two temporospatial factors: a centrally distributed factor differing between correct and incorrect responses and a more frontoparietally distributed factor contributing to both ERN/Ne and CRN. These data support the notion that ERN/Ne and CRN might reflect a combination of two underlying processes: an error-sensitive and an outcome-independent aspect of response monitoring. Descriptors: Performance monitoring, Error processing, Error-related negativity (ERN), Correct-related negativity (CRN), Error positivity (Pe) Performance monitoring is a necessary prerequisite for behavioral adaptation and flexibility. Over the past 20 years, research has focused on brain processes related to error commission and its monitoring. In the event-related electroencephalogram (EEG), the error negativity (Ne) or error-related negativity (ERN) is observed shortly after erroneous responses in reaction time tasks (Falken- stein, Hohnsbein, Hoormann, & Blanke, 1990; Gehring, Goss, Coles, Meyer, & Donchin, 1993). The specificity of the ERN/Ne for errors has been challenged by the observation of a negative deflection following correct responses (correct-related negativity: CRN; Coles, Scheffers, & Holroyd, 2001; Ford, 1999; Vidal, Burle, Bonnet, Grapperon, & Hasbroucq, 2003; Vidal, Hasbroucq, Grap- peron, & Bonnet, 2000). The issue of one or more processes under- lying ERN/Ne and CRN has important implications for models of performance monitoring. In addition to the theoretical implica- tions, studying the functional significance of response-related com- ponents might also help to understand deficits in action control and behavior adaptation observed in psychopathological and neurologi- cal conditions (Taylor, Stern, & Gehring, 2007; Ullsperger, 2006). The present study was designed to further investigate functional characteristics of response-related negativities and their underlying processes. The ERN/Ne is a negative event-related potential (ERP) com- ponent over frontocentral scalp locations appearing immediately after the commission of an erroneous response. Source localization and simultaneous EEG and fMRI recordings point to the medial frontal cortex or, more specifically, to the rostral cingulate zone (RCZ) as a major source (Debener et al., 2005; Dehaene, Posner, & Tucker, 1994). Originally, the ERN/Ne was considered to reflect an error detection process resulting from a comparison between the actual and the required response (Falkenstein et al., 1990; Gehring et al., 1993). This view was extended by the reinforcement learning theory that considers the ERN/Ne as a reinforcement learning signal generated by the anterior cingulate cortex (ACC), which, in turn, is driven by midbrain dopamine signals indicating “worse than expected” outcomes (Holroyd & Coles, 2002). An alternative model, the conflict theory, posits that the ERN/Ne is associated with conflict monitoring in the ACC that arises from multiple simultaneously active response tendencies (Botvinick, Braver, Barch, Carter, & Cohen, 2001;Yeung, Botvinick, & Cohen, 2004). Further models regard the ERN/Ne as a neural indicator of error likelihood or expected risk (Brown & Braver, 2005, 2008). Despite unresolved issues about the underlying processes reflected by the ERN/Ne, the mentioned models converge on the assumption that the ERN/Ne triggers the adjustment of cognitive control serving to prevent future errors (Ridderinkhof, Ullsperger, Crone, & Nieu- wenhuis, 2004). This interpretation is supported by findings of positive correlations between ERN/Ne amplitudes and post-error slowing (Debener et al., 2005; Gehring et al., 1993). Partly similar results were found for the error positivity (Pe), an ERP that peaks This work was supported by the German Research Foundation EN 906/1-1. Address correspondence to: Tanja Endrass, Ph.D., Humboldt- Universität zu Berlin, Department of Psychology, Rudower Chaussee 18, 12489 Berlin, Germany. E-mail: endrass@gmail.com Psychophysiology, •• (2012), ••–••. Wiley Periodicals, Inc. Printed in the USA. Copyright © 2012 Society for Psychophysiological Research DOI: 10.1111/j.1469-8986.2012.01365.x 1