Image from EyeInfo
Background
Even in an era of increasing technology dependence, reading remains an essential daily task. It is critical in education and the work place, where knowledge is transmitted most often through text. In fact, a child’s future economic success can be predicted by their reading proficiency as early as the fourth grade, and there is clear indication that students from low-income families have poorer reading abilities (1). This shows a clear need for a cost-effective solution capable of monitoring how children are reading, in order to provide schools and caregivers the information they need to help a child develop into a successful reader.
Reading occurs via a combination of fixations and saccades. Through the positional measurements provided by an eye tracker, it is possible to determine a reader’s fixation and saccade data. A variety of inferences can be drawn from these metrics, including a user’s gaze pattern or behavior and the complexity of the material being read. Many of these inferences apply directly to reading skills, and thus can provide feedback on how an individual reads.
Reading occurs via a combination of fixations and saccades. Through the positional measurements provided by an eye tracker, it is possible to determine a reader’s fixation and saccade data. A variety of inferences can be drawn from these metrics, including a user’s gaze pattern or behavior and the complexity of the material being read. Many of these inferences apply directly to reading skills, and thus can provide feedback on how an individual reads.
Scope
This project aims to create an affordable consumer-friendly software toolset that will work in conjunction with existing gaze tracking hardware to quantify and evaluate reading patterns as well as store user session histories. Using experimental metrics, it will provide instant feedback to the user outlining their reading performance. Reading performance will be defined by a user’s reading speed, a breakdown of document sections skipped or skimmed, possible indications of distraction or boredom, and a list of sections or vocabulary words where the user somehow struggled (as indicated by prolonged fixation and redundant reading). This feedback will be presented in an intuitive package with the potential for a user to compare results to previous sessions or global averages. A user’s progress history will be stored for recall at a later point in time, and potential use in big data analysis. This project will not include the design of any hardware, as there already exist eye-tracking devices on the market. The project will not serve to replace any existing clinical-grade diagnostics.