Predicting student performance in an augmented reality learning environment using eye-tracking data
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
This paper investigates the use of eye-tracking data as a predictor of student performance in an augmented reality (AR) learning environment. 33 undergraduate students enrolled in an ergonomics course at the University of Missouri-Columbia participated in an AR biomechanics lecture consisting of 14 modules. Following each module students answered learning comprehension questions to test their understanding of the lecture material. An additional dataset was recorded for each module in which the participant perfectly follows the virtual instructor throughout the learning space. This dataset, referred to as the baseline, can be used as a comparison tool to gauge how well students follows the lecture material. Two methods are proposed to quantify the student's attention level for each module. The average difference method calculates the average distance between the student and baseline coordinates for each module. The distraction rate method expands upon the average difference method and aims to reduce the amount noise detected. This is done by incorporating a minimum distance threshold, a binary detection signal, and a moving average window. Both metrics are tested as factors in a set of logistic regression models to determine whether they can accurately predict student answer correctness. Average difference showed a correlation with student answer correctness, but with an underwhelming level of significance. Distraction rate outperformed average difference and proved to be a strong and statistically significant predictor of student answer correctness. Finally, two feedback systems are proposed which use distraction rate to detect when students have become distracted so that their attention can be regained through the use of module-based feedback or a real-time attention guidance system.
Table of Contents
DOI
PubMed ID
Degree
M.S.