We designed a browser-based screener, Dytective, to detect risk of dyslexia across different alphabetic languages. Dytective consists of a variety of linguistic exercises based on analysis of a large corpus of errors made by people with dyslexia. Results of two studies of Dytective, one with English-speaking children and the other with Spanish-speaking children, demonstrated significant differences on measures of rate and number of errors between children with and without diagnosed dyslexia for both languages. Using the data collected in these studies, we implemented a statistical model that is able to predict whether or not a child has a diagnosis of dyslexia with 85.85% accuracy.
Special Faculty, Carnegie Mellon University & Founder, Change Dyslexia,
HCI Institute, Carnegie Mellon University
You must be logged in and own this session in order to