Exploring Learning Behaviour of Students with Special Needs through Handwriting Logs

Abstract

Students' handwriting log data can now be collected through digital devices. Several studies have utilized this data to understand the condition of learners, but few have specifically examined the writing behavior of students with special needs while working on task-based activities. This study aims to identify characteristics of the writing behavior of students with special needs through the analysis of handwriting log data. Four main features were analysed: the number of strokes, the number of erases, the number of pauses, and the duration of writing. The agglomerative hierarchical clustering algorithm was used to group students based on the similarities in their behavior. One group showed higher average values across all features, while the other showed lower values. These findings suggest that students with special needs have diverse learning patterns, which can be mapped through handwriting data and possibly used to support more adaptive learning design.