본문 바로가기
  • Home

The Educational Effects of AI Feedback in Digital Writing: A Case Study at A University

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2025, 11(2), pp.15~22
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : March 12, 2025
  • Accepted : April 3, 2025
  • Published : April 30, 2025

Park, Hye Rim 1 Lee Uiyeon 2

1배화여자대학교
2고려대학교

Accredited

ABSTRACT

This study aimed to examine the effects of an AI feedback–based digital writing program. Specifically, we statistically verified changes in participants’ digital writing competencies in an AI-based extracurricular program at A University and conducted an in-depth exploration of AI feedback experiences through reflective journal analysis. Over a seven-week period, learners drafted texts based on weekly prompts and repeatedly revised them using immediate AI feedback. The results showed significant improvements in areas such as digital information searching and utilization, writing planning and organization, and document revision. Notably, learners’ writing self-efficacy and interest increased through the gradual refinement of their work. However, their beliefs about digital writing did not undergo major changes in the short term, and limitations arose from AI feedback’s insufficient consideration of creative and emotional aspects, as well as contextual nuances. These findings suggest the importance of maximizing the benefits of AI feedback while also supplementing it with instructor or peer feedback to address contextual and affective elements. Furthermore, instructional designs should train learners to selectively accept or reject AI feedback according to writing genre and individual characteristics. This study holds significance in empirically demonstrating that AI-based writing instruction can effectively enhance learners’ writing performance and motivation in actual educational settings.

Citation status

* References for papers published after 2023 are currently being built.