Enhancing Writing Assessment With Automated Tools

What is it about?

This paper introduces ScriptSense, an automated linguistic analysis platform developed to examine children's writing through measures of morphology, syntax, and lexical diversity. Since the completion of this study, the project has evolved substantially. ScriptSense now serves as the foundation of ScriptToolKit, a unified platform released in May 2026 that integrates transcript preprocessing, linguistic analysis, profile generation, and automated reporting into a single system. As a result, this publication reflects both an evaluation of the original platform and an important step in the continued development of computational tools designed to support research and clinical practice in speech-language pathology and literacy.

Why is it important?

This work is important because it shifts the conversation about writing assessment from what students do wrong to what their language reveals about development. Traditional writing assessments often focus on correctness, conventions, and overall quality scores, but those measures provide limited information about the underlying linguistic knowledge students are using when they write. By examining morphology, vocabulary, and syntax simultaneously, ScriptSense demonstrated how automated tools can make these linguistic patterns more visible and accessible to researchers, educators, and speech-language pathologists. The study also represents an early step toward leveraging computational linguistics within communication sciences and disorders. As language samples become increasingly recognized as valuable sources of information, there is a growing need for tools that can efficiently analyze large amounts of text while preserving meaningful linguistic detail. This work contributes to that goal by illustrating how automated analyses can complement, rather than replace, professional judgment and help generate richer descriptions of children's written language.