AI-POWERED FEEDBACK SYSTEM FOR WRITING: A REVIEW OF AUTOMATED WRITING EVALUATION (AWE) TOOLS

Abstract

This systematic literature review aims to explore how AI-powered feedback tools contribute in learners’ writing. The researcher analyzed the results from 50 empirical studies that published from 2019 to 2024. The studies were selected via database searches, defined inclusion criteria and thematic analysis. The findings indicate that Automated Writing Evaluation (AWE) tools such as Grammarly, Quillbot and DeepL Write significantly improve the writing precision, fluency and grammatical proficiency through immediate and customized feedback. Additionally, the review examines that AWE tools can enhance the organization of writing, overall writing quality, while also encouraging self-directed learning. However, the review pointed out that learners relying too much on AWE and the fact that it could lower their critical thinking and creativity. More challenges like sometimes AWE might provide limited feedback depth and concerns about accuracy in evaluating complex writing aspects remain significant. This article highlights the pedagogical implications of Integrating AI-powered feedback and suggests directions for future research, particularly in balancing human and automated feedback to optimize writing skill.