DysMusic: detecting dyslexia by web-based games with music elements
Maria Rauschenberger · 2016 · Proceedings of the 13th International Web for All Conference (W4A) · doi:10.1145/2899475.2899503
Summary
This doctoral consortium paper proposes DysMusic, a web-based serious game that uses music elements to detect dyslexia in children aged 3 to 6—before they enter formal schooling and begin experiencing academic failure. The research is motivated by longitudinal evidence showing that children diagnosed with dyslexia at age 8 go on to achieve lower school performance and higher unemployment rates by age 25, making early detection critical. The approach builds on the connection between dyslexia and auditory/temporal processing difficulties: since people with dyslexia often have short-term memory challenges, music-based tasks like "Which sound did you hear first?" or "Which sound is pitched higher?" could serve as indicators. The key advantage of using music rather than language-based tasks is the potential for language independence—unlike linguistic exercises, music-based detection could work across different languages. The research follows a Human-Centered Design process and plans user studies with schools and dyslexia associations in Germany and Spain. The system would extend the existing Dytective platform, a multilingual web application that has already demonstrated the feasibility of detecting dyslexia through online games analyzing word errors and interaction patterns.
Key findings
The author reports preliminary results in two areas. For intervention, over 1,000 spelling errors written by German children with dyslexia were collected from schools, classified by phonetic and visual features, and used to create 2,500+ word exercises integrated into an iOS game—with evidence that error-based computer game exercises significantly improve spelling skills of children with dyslexia in Spanish. For detection, a set of linguistic exercises was created for a German version of Dytective, with initial results confirming that web-based exercises can detect dyslexia; a study with 30 participants was underway to determine accuracy for German. The proposed DysMusic game would combine these linguistic elements with music-based tasks, using an adapted memory game format where children find matching sounds behind digital cards rather than matching pictures. Dependent variables would include mouse tracking data, performance measures (hits, clicks, scores), and subjective questionnaire responses. The second phase would optimize the diagnosis threshold by enriching the existing Dytective prediction algorithm with music-based features.
Relevance
This research addresses the important problem of early dyslexia detection, which can dramatically alter life outcomes by enabling intervention before school failure erodes a child's confidence and academic trajectory. The music-based approach is novel and has two practical advantages over traditional screening: it can be used with pre-literate children (ages 3-6) who cannot yet be assessed through reading and writing tasks, and its potential language independence could make it viable for multilingual and cross-cultural deployment. The web-based delivery model makes screening accessible and cost-effective at scale, compared to expensive clinical assessments. However, as a doctoral consortium paper, this work was at an early conceptual stage—the music-based detection component had not yet been implemented or tested, and the theoretical connection between music perception tasks and dyslexia indicators had not been empirically validated. The research builds on the same group's work at Universitat Pompeu Fabra (including Luz Rello's Dytective), which has established credibility in game-based dyslexia research.
Tags: dyslexia · dyslexia screening · serious games · gamification · music accessibility · early detection · web-based assessment · learning disabilities