Identifying Significant Indicators of Eye-movement and EEG-based Attention to Predict Reading Performance

Abstract

It is important to extract students’ reading data to predict their reading performance. This study aims to identify significant indicators of eye-movement and EEG-based attention and to test their predictive effectiveness on reading performance. Data were collected from 56 undergraduate students who read an illustrated science text about geography. Out of 21 reading indicators, 16 were found to have a significant correlation with reading performance. The multiple regression model suggested that Whole time, Text-diagram, Test-text, Medium attention, and High attention were significant indicators. They predicted 62.5% of the variation in students’ reading performance.