Fostering a ‘Human with AI’ approach for evaluating students’ writing in English
Abdu M. Talib Al-kadi · Studies in Linguistics Culture and FLT · 2025
With the increasing interest in Artificial Intelligence (AI) within academia, Automated Writing Evaluation (AWE) has gained significant traction. However, its full impact on writing skills remains a topic of contention. This paper examines how AI-based feedback helps undergraduate students overcome challenges they encounter when they write essays as part of their university studies. A text-oriented approach to writing evaluation was adopted to demonstrate how Grammarly was utilized to screen and score a sample of 22 short essays (11,050 words). In addition to assigning numerical scores to the essays, Grammarly provided student writers with detailed feedback, including error reports that were subsequently analyzed. Then, two focus group discussions were administered to shed more light on Grammarly-based evaluation. Findings showed that AWE encourages learners to observe their errors and refine their essays accordingly before submitting them to teachers for scoring. Nevertheless, such a tool per se is short to provide a thorough evaluation. It could be used in tandem with peer review and teachers’ evaluation. The paper closes on some implications and suggestions to foreground AWE in academic writing courses besides, but not a surrogate to, human raters’ feedback.