Evaluating the Validity of AI-Generated Writing Feedback for Students With Learning Disabilities
Students with learning disabilities need clear, direct, and useful writing feedback. However, teachers often find it hard to provide this consistently because of time and workload pressures. New developments in generative artificial intelligence (AI) might help improve both the frequency and quality of feedback for students. Still, we know very little about whether AI-generated comments are valid or useful for struggling writers. This qualitative study looked at teachers' views on writing feedback from the AI-SCORE automated writing evaluation tool. Nine upper elementary and middle school teachers took part in two focus groups. They reviewed AI-generated feedback on student informational writing. They also discussed its clarity, teaching value, and suitability for students with disabilities. Thematic analysis revealed several areas for improving AI-generated feedback. Teachers want more guidance on specific revisions, language that suits students’ development, a consistent tone, and alignment with writing goals. They shared cautious optimism, noting that AI feedback could enhance teaching capacity and provide more practice opportunities for students, if it meets their needs. The findings stress the importance of teacher involvement, the potential benefits of AI for writing instruction, and key factors for creating AI tools that effectively support students with learning disabilities.
Authors
- Sean J. Smith (ORCID: https://orcid.org/0000-0002-4566-8018)
- Adam Carreon (ORCID: https://orcid.org/0000-0001-9677-8025)
- Samantha R. Goldman (ORCID: https://orcid.org/0000-0003-0975-256X)
Institutions
- University of Kansas (US)
- Georgia Southern University (US)
Publication Details
- Journal
- Journal of Special Education Technology
- Published
- 2026-11-02
- DOI
- https://doi.org/10.1177/01626434261498189
- Primary Topic
- Artificial Intelligence in Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00