Expanding the spectrum of dysgraphia: a data-driven strategy for estimating handwriting quality

Thibault Asselborn, Mateo Chapatte & Pierre Dillenbourg

This paper proposes new ways of assessing handwriting, a critical skill in any child’s educational journey. Traditionally, a pen and paper test, called the BHK test (Concise Evaluation Scale for Children’s Handwriting), is used to evaluate children’s handwriting in French-speaking countries. Any child with a BHK score above a certain level is diagnosed with dysgraphia, meaning that they are then eligible for financial coverage for therapeutic support. We previously developed a version of the BHK for tablets which provides strong data on the dynamics of writing (acceleration, pressure, and so forth). The principal model was tested with dysgraphic and non-dysgraphic children. With this contribution, we differ from the original BHK test due to three reasons.

First, in this case, we are not interested in a binary result but rather a result on a scale of handwriting difficulties, from the mildest cases to the most severe. Therefore, we want to calculate the difference between a child’s score and the average score of a child of the same age and gender.

Second, our model analyses dynamic characteristics that are not accessible on paper; the BHK is therefore useful in this instance. Using the PCA (Principal Component Analysis) reduced the 53 handwriting characteristics to three dimensions, independent of the BHK. Additionally, we have verified that when we group our data set according to any of these three aspects, we can accurately detect dysgraphia in children.

Third, dysgraphia is a general concept that encompasses a wide variety of writing difficulties. Two children with the same overall score may have totally different types of writing difficulties. For example, one child may apply uneven pen pressure, while another may have trouble controlling their writing speed.

Our new test not only provides an overall score, but also includes four specific scores for kinematics, pressure, pen tilt and static characteristics (letter shape). Replacing an overall score with a more detailed profile allows the selection of very specific remediation sets for each profile.