Optimizing the usability of reading assessments with eye-tracking on a mobile device

Abstract: An estimated 20% of Swedish schoolchildren suffer from some form of reading difficulty such as dyslexia. Late diagnosis of dyslexia leads to consequences for both the individual and society. To identify such reading difficulties, over 50 different forms of reading assessments are being used in schools around Sweden. By using eye-tracking technology, one could minimize the time for these assessments and make them less prone to external biases. In this study such a solution is evaluated with regards to usability. Using widely available technology, mobile devices with built-in front-facing cameras, a reading assessment with integrated eye-tracking was developed. The aim of this study was to evaluate the usability of the solution and find areas to be improved. This was done through allowing 14 participants to use the application and answer a usability survey based on the System Usability Scale (SUS). The results showed an average SUS-score of 87.0, equivalent to a grade of A+, which reflected the positive responses in the open questions of the survey. The main area of improvement found was the speed of the calibration process, as this solution was found too slow. To further improve this solution, head and hand movement must be taken into consideration by the eye-tracker to allow free hand-held usage with less supervision. Further improvements such as larger screens and added machine learning are also suggested. Conclusively the results of this study show promise if such improvements are accomplished.