Digital Characterization of Primary School Pupils' Handwriting with a Sensorized Ink Pen

Difficulties in mastering handwriting affect 10-30% of pupils, with detrimental effects if untreated. Thus, their early detection is of utter importance. Technology can enrich the standard pen-and-paper assessment with quantitative information on the handwriting process, provided it can describe both its development and impairments in childhood. In this work, 530 pupils from 1^st to 5^th grade performed a subtest from the Italian clinical gold standard for writing fluency assessment. They wrote the sequence of numbers as words, as fast as possible, for one minute in uppercase font. An innovative sensorized ink pen (SIP), able to collect kinematic and dynamic signals while writing on paper, was used for data acquisition. From the data, temporal, fluency, and pressure indicators were computed and subjected to statistical analysis to evaluate handwriting learning curves and sensitivity to difficulties, revealed by the writing fluency test. All domains showed a significant learning effect (p-value < 0.001). A relevant evolution in the handwriting pattern in the first three years emerged, with significant differences between adjacent classes (p-value < 0.001), followed by a plateau. The indicators were also statistically different (p-value < 0.001) between proficient and nonproficient writers. Beginner and nonproficient writers were characterized by reduced speed both with the SIP on paper and in air, jerky movements and scarce modulation of the force exerted on the writing surface. These findings, in line with the existing literature, foster the integration of the SIP in the early screening of handwriting difficulties.