Learner agency in revising GenAI ‐generated statements of purpose

Yuanlan Jiang et al.

British Journal of Educational Technology2026https://doi.org/10.1111/bjet.70041article
AJG 2ABDC A
Weight
0.37

Abstract

This study investigates how English as a foreign language (EFL) learners enact their agency in revising generative AI (GenAI)‐generated academic texts, in this case the statement of purpose, or SOP. Conducted at a key university in Southern China, the study involved 121 English‐major students. This study included two phases. First, we used the Curriculum Vitae (CVs) submitted by the participants to generate SOP drafts through a standardized prompt in ChatGPT‐4. Second, the participants acted as evaluators and editors, engaging with the GenAI‐generated SOPs to varying extents. Nine of the participants attended semi‐structured retrospective interviews. Textual comparison of original and final SOPs was conducted to identify patterns of agency, non‐parametric analysis was performed between learners' ratings and their identified agentic patterns, and thematic analysis was applied to the interview data. The study revealed three agentic revision patterns: compliance‐oriented acceptance, form‐oriented modification, and content‐oriented innovation. Differences in learners' ratings of GenAI‐generated SOPs were found across the three patterns. Two factors influenced these patterns: authentic voice versus GenAI‐constructed textual voice and enthusiasm to present intentions. The study suggests that empowering EFL learners to activate their agency in using GenAI tools enhances their academic writing and supports them in meeting specific task requirements. Practitioner notes What is already known about this topic EFL learners often face challenges in academic writing, particularly in incorporating GenAI‐generated material effectively. Agency in revision processes is important for EFL learners using GenAI tools. Previous research highlights the potential of GenAI in supporting writing tasks, but there is limited insight into how EFL learners express their agency during revisions of GenAI‐generated content. What this paper adds This study explored how EFL learners enact their agency when revising GenAI‐generated academic texts (ie, SOPs). We identified three specific patterns of agency enactment during the revision process, namely, compliance‐oriented acceptance, form‐oriented modification, and content‐oriented innovation. The research highlighted factors influencing these agency patterns, including authentic voice versus GenAI‐constructed textual voice and enthusiasm to present intentions. Implications for practice and/or policy Educators should promote learner agency by encouraging students to recognize and manage their revision strategies, including acceptance, modification, and innovation while using GenAI tools, through instructional strategies that empower effective agency. EFL writing instructors should develop critical thinking skills and encourage learners to develop their unique authorial voice, particularly when revising personal and complex texts (e.g., SOPs) to avoid overreliance on GenAI. Teachers should guide learners in balancing their authentic identities with GenAI enhancements, ensuring students remain the agents in shaping their voices and conveying the emotional resonance of their experiences in their SOPs.

1 citation

Open via your library →

Cite this paper

https://doi.org/https://doi.org/10.1111/bjet.70041

Or copy a formatted citation

@article{yuanlan2026,
  title        = {{Learner agency in revising GenAI ‐generated statements of purpose}},
  author       = {Yuanlan Jiang et al.},
  journal      = {British Journal of Educational Technology},
  year         = {2026},
  doi          = {https://doi.org/https://doi.org/10.1111/bjet.70041},
}

Paste directly into BibTeX, Zotero, or your reference manager.

Flag this paper

Learner agency in revising GenAI ‐generated statements of purpose

Flags are reviewed by the Arbiter methodology team within 5 business days.


Evidence weight

0.37

Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40

F · citation impact0.16 × 0.4 = 0.06
M · momentum0.53 × 0.15 = 0.08
V · venue signal0.50 × 0.05 = 0.03
R · text relevance †0.50 × 0.4 = 0.20

† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.