To evaluate the learning attention and effectiveness in three remote learning approaches using EEG, eyetracking and traditional exam

In past three years, many schools have switched to remote learning in response to the new pneumonia epidemic, but remote learning may be less effective due to unsupervised learning or the surrounding environment. This study investigated the impact of three remote learning methods, including (1) 2D video without face (slide + teacher's lecture); (2) 2D video with face (slide + teacher's lecture + showing face); and (3) immersive virtual reality (VR) using 360-degree video. Total 17 students were recruited in our experiment. Based on the grade point average (GPA) in last semester, all participants were divided into two groups: higher GPA and lower GPA first, and then subdivided into three groups with different remote learning method. In order to realize the degree of attention for each participant during the experiment, electroencephalography (EEG) and eye tracking data were collected simultaneously. Moreover, the exam after experiment was also applied to each participant for realizing the learning effectiveness. The results showed that no significant difference exists in the the exam score after watching video between three remote learning methods. The reason might be that the discrimination index of exam were less to distinguish the learning effectiveness. From the EEG scores, the VR group shows higher attention level. However, the VR group shows the worst eye tracking attention score. It might be caused by that the VR content is relatively rich than 2D video. Moreover, for the EEG analysis, we used two different machine learning models for the attention classification, and the final results showed that logistic regression had better classification results. In the future, we would like to use EEG combined with VR for real-time attention alerts to give immediate alerts when students are not concentrating in order to obtain better learning results.