Effects of a ChatGPT-Assisted Writing Intervention on the Productive Writing Skills of Students With Learning Disabilities: A Single-Case Experimental Design Study

Effects of a ChatGPT-Assisted Writing Intervention on the Productive Writing Skills of Students With Learning Disabilities: A Single-Case Experimental Design Study

Authors

  • Susanne Hoff
  • Matthias Grünke
  • Janine Bracht
  • Anne Barwasser

Keywords:

ChatGPT, self-regulated strategy development, learning disabilities, AI-generated feedback, inclusive writing instruction, single-case design, textuality

Abstract

This single-case design study explored the impact of AI-supported writing instruction for students with learning disabilities (LD). We examined how ChatGPT-generated feedback embedded within a structured Self-Regulated Strategy Development (SRSD)-based writing intervention affected students‘ writing productivity, as measured by Total Words Written (TWW), and text quality, as measured by the Teacher Evaluation of Story Elements rubric (TESE). Drawing on the Hayes and Flower (1980) writing process model and the SRSD framework, the intervention provided scaffolded task-level feedback using standardized prompts and iterative revision cycles. The effects were evaluated across alternating baseline, intervention, and follow-up phases using visual inspection and non-parametric overlap indices. Results indicated positive effects on both TWW and TESE, with consistent level effects across intervention phases. Moreover, students gradually internalized feedback routines, indicating the development of metacognitive strategies. The findings highlight the pedagogical potential and limitations of AI-mediated scaffolding in inclusive education and call for critical reflection on its implementation.

Downloads

Published

2026-04-02

Issue

Section

Article
Loading...