Valence-dependent brain potentials of processing augmented feedback in learning a complex arm movement sequence

ERPs in the EEG were scrutinized in learning a complex arm movement sequence with the aim to examine valence effects on processing augmented feedback during practice. Twenty-four healthy subjects practiced one session with 192 feedback trials according to an adaptive bandwidth feedback approach with a high informational level of feedback information (i.e., amplitude and direction of errors). The bandwidth for successful performance (increase of a score for a monetary competition) was manipulated to yield a success rate (positive feedback frequency) of approximately 50% adaptive to the current performance level. This allowed a variation of feedback valence unconfounded by success rate. In line with our hypotheses, the EEG data showed a valence-dependent feedback-related negativity (FRN) and a later fronto-central component at the FCz electrode as well as a P300 component at the Pz electrode. Moreover, the P300 and amplitudes in the FRN time window reduced in the second half of practice but were still dependent on feedback valence. Behavioral adjustments were larger after feedback with negative valence and were predicted by the late fronto-central component. The data support the assumption of feedback valence-dependent modulation of attentional cognitive involvement in motor control and learning.