This pilot study investigates the feasibility of using electroencephalogram (EEG) signals to assess students’ attention in a multimedia learning environment and to examine whether EEG-derived features can differentiate correct and incorrect responses to comprehension questions. EEG data were collected from 25 university students while they watched a short educational video and answered two follow-up questions. Power spectral density values of Delta, Theta, Alpha, Beta, and Gamma frequency bands were extracted and used as input features for five classification algorithms: Bayesian Classification, Logistic Regression, C5.0, CHAID, and Artificial Neural Networks. Model performance was evaluated using accuracy, sensitivity, specificity, and F1-score. Bayesian Classification achieved the highest overall performance for both questions. Across models, Beta and Gamma band activities emerged as the most informative features for distinguishing correct from incorrect responses. These findings indicate that EEG-based measures can provide objective indicators of attention-related cognitive engagement in multimedia learning contexts. The results demonstrate the potential of combining EEG and machine learning for attention assessment and support the feasibility of further large-scale investigations.
Karabakla et al. (Wed,) studied this question.