Analyzing Eye Gaze of Users with Learning Disability

Abstract

This paper investigates eye gaze movements of children with different reading abilities due to learning disability. We analyzed eye gaze fixations using cluster analysis and velocity-based fixation classification algorithms. We used a non-invasive, screen mounted, low-cost eye tracker to collect data from 30 participants including students with learning disabilities. We analyzed raw gaze points, fixations and saccades, and noted that the density of gaze points around fixation. Total number of fixations and regressions (backward gaze movement) were statistically significantly correlated to the reading ability of students. We concluded that by analyzing different eye movements for a representative reading task, we can detect early signs of learning disabilities such as dyslexia.