A proof-of-concept EEG study of responses to peer speech in collaborative learning

Source abstract

Background This study focused on internal aspects, including cognitive states and motivational engagement, that may be overlooked by external indicators of participation in collaborative learning, such as utterance counts or observable behaviors. This proof-of-concept feasibility study explored whether neurophysiological responses related to auditory attention could provide information relevant to interpreting conventional behavioral indicators of participation. Using electroencephalogram (EEG) indices (N1 and P2 components), associations with self-regulated learning and interest during collaborative learning were explored. Methods EEG data were collected during real, discussion-based collaborative learning activities conducted in a university classroom setting. EEG data were recorded using wearable devices from nine Japanese university students enrolled in the course. Utterances were categorized as surface- or deep-level regulation utterances according to definitions from a prior study. Using a multivariate temporal response function approach, the N1 and P2 amplitudes elicited by each type of utterance were estimated. The degree of self-regulated learning was indexed by the number of deep-level regulation utterances, and interest in learning was evaluated using subjective rating scores. The relationships between these indices and the N1 and P2 amplitudes were analyzed using linear mixed-effects models. Results The number of deep-level regulation utterances was significantly associated with the N1 amplitude, irrespective of whether the utterances were categorized as deep or surface regulation. In contrast, subjective interest ratings were significantly associated with the N1 amplitude, particularly in response to deep-level-regulation utterances. Conclusion The N1 component, which is related to auditory attention, showed exploratory associations with self-regulatory behavior and learners’ interest.