Learners' engagement with AI-automated, peer, teacher feedback: A three-pronged approach to improving academic writing

This presentation reports on learners’ engagement with three types of feedback in their revision of their essays and discusses how their engagement affected the quality of their essays.

The importance of feedback in learning is widely acknowledged in the scholarship of teaching and learning and is strongly felt in classroom practices. Feedback is a powerful pedagogical tool to support students’ learning (Poulos & Mahony, 2008; Yang et al., 2021) as it scaffolds students to notice the gaps in their knowledge and support them in growing beyond their current level. Constructive feedback is frequently found to correlate positively with students’ academic performance (e.g., Burns et al., 2021), motivation (e.g., Fang et al., 2021), and engagement in learning (e.g., Carvalho et al, 2021).

While feedback is a powerful pedagogical tool to help learners develop their language skills, it is their engagement with feedback that will help them in their development (Zhang and Hyland, 2022). Hence, in this study, we followed Zhang and Hyland’s design in providing our learners with a three-pronged approach to feedback (hence three rounds of feedback and revisions) to help them develop their academic writing skills: (1) AI-automated feedback, (2) peer feedback, (3) teacher feedback. AI-automated feedback targeted areas such as rhetorical moves and genre-specific suggestions for improvement, while peer feedback was designed to focus on aspects such as clarity of meaning, development of arguments, author’s tone and voice. Teacher feedback, as the last phase in this process, focused on accuracy and depth of critical thinking in their arguments, stylistic choice of lexicons, etc.

The data was collected from year 1 engineering students enrolled in a core module in communication skills where they learned, among other skills, to produce academic essays. The collected data included four drafts of their essays and the feedback they received in each of the revision rounds. We traced the changes they made in response to the feedback and looked at how they incorporated it into their revisions. Subsequently, we compared the changes in the quality of their essays, both holistically and structurally. The results showed that the revision in response to teacher feedback improved the essay quality more significantly than the other types of revision, highlighting again the importance of human teacher feedback in the age of AI.

In this presentation, I will report on the details of students’ engagement with the feedback, give examples of their revisions and share students’ perceptions of each feedback round’s effectiveness. Additionally, I will discuss the implications of these findings for educators and policymakers. By understanding how different types of feedback influence student engagement and learning outcomes, we can better design instructional approaches that maximize the benefits of feedback. Furthermore, the integration of AI in educational settings poses both opportunities and challenges, which will be explored in the context of providing meaningful and effective feedback to learners. This comprehensive analysis aims to offer valuable insights and practical recommendations for enhancing academic writing instruction through a balanced and integrated feedback approach.