Meta-Analysis of Using AI-Based Feedback Systems in Developing College Students’ Academic Writing Skills

In a technology-driven world where using artificial intelligence (AI) systems have dominated in all fields, educational institutions have become keen to embrace innovative tools and applications as a fundamental part of teaching-learning to promote learners’ skills and competencies. Automated writing feedback systems, which are AI-based tools, are being used excessively at universities to provide students with immediate and timely written feedback; hence, reviewing, analyzing, and synthesizing the methods and results of existing studies in this regard can be considered a crucial headway. Thus, the current study aimed at meta-analyzing the previous studies relatable to the usage of AI-based feedback systems to improve the academic writing skills of university students. Consequently, 35 studies published between 2015 and 2023 were analytically summarized and grouped based on the meta-analysis guidelines established by Cooper (Management Decision 36:493–502, 1998). The results revealed numerous benefits and some challenges of utilizing this technology. It was displayed, for example, that using AI-based feedback systems can significantly improve students’ academic writing skills, increase their engagement, promote self-regulation, trigger meta-cognitive writing skills, and provide personalized feedback that facilitates the learning process. Conversely, some challenges were highlighted, especially the overuse of such technologies that can lead to dependent and reliant learners, in addition to the failure of these systems to provide feedback on writings that necessitate higher-order thinking skills, meanwhile offering generic feedback that ignores individual differences. Accordingly, a model for the optimal use of such systems was suggested in the conclusions.