A composite methodology for supporting early-detection of handwriting dysgraphia via big data analysis techniques

Abstract

Handwriting difficulties represent a common cause of under-achievement in children’s education and low self-esteem in daily life. Since proper handwriting teaching methods can reduce dysgraphia problems, the evaluation of these methods represents an important task. In this paper a methodology to compare visual and spatio-temporal teaching methods is proposed and applied in order to assess the influence of different teaching approaches on handwriting performance, via big data analysis techniques. Data was collected from children in their final years of primary school, when cursive writing skills have typically been mastered. Qualitative and kinematic parameters were considered: the former were calculated by means of quality checklists, whereas the latter were automatically extracted from digitizing tablet acquisitions. Results showed significant differences in pupils’ handwriting depending on the teaching method applied.