Singapore Institute of Technology
Browse

Leveraging AI in the Post-Task Phase: Automated Feedback and Focus on Form in Academic Writing

presentation
posted on 2025-05-09, 02:49 authored by R RosmawatiR Rosmawati

Abstract:

This presentation reports on the results of an action research project that leveraged on AI in the post-task phase to help students improve their academic writing skill in a TBLT class for first-year engineering students.


Based on teachers’ observation and the module feedback in the previous years, it is evident that the students frequently struggle with the rhetoric of a thesis-driven essay, particularly in conceiving a thesis statement and supporting it through logical arguments that demonstrate substantial evidence of critical thinking. Therefore, in this project, we leveraged on an automated writing evaluation tool, called AcaWriter, in the first stage of our post-task phase to help students focus on form (i.e., on the rhetorical moves in their essay.). This AI tool was chosen because it can analyse the rhetorical moves in students’ texts and generate genre-specific feedback. The second and third stages of our post-task phase included peer-review (targeting meaning clarity) and teacher consultation (targeting the depth of arguments). We designed this 3-stage post-task phase following Zhang and Hyland’s (2022) recommendation and called it an integrated approach to formative feedback. Students revised their drafts after each stage in the post-task phase and submitted the revision to the assessment dropbox, although only the final draft was marked for assessment.


We traced the changes in the drafts and compared them to the feedback in each of the stages. Our results show that our students benefitted from the post-task phase in terms of their increased (and more explicit) use of genre-appropriate rhetorical moves to express their authorial voice in both the thesis statement and the subsequent supporting arguments. In this presentation, we will show the students’ interaction with the AI feedback, as well as the other two types of feedback, and demonstrate how such focus on form helped them improve their academic writing skills.

Funding

ALIGN Grant

History

Journal/Conference/Book title

11th International Conference on Task-Based Language Teaching

Publication date

2025-04

Version

  • Published

Corresponding author

rosmawati@singaporetech.edu.sg

Project ID

  • 14859 (P-ALI-A203-0018-OOE) An Integrated Approach to Feedback: Automated, Peer, and Teacher Feedback for Improvement in Writing