From Prompt to Argument: A Qualitative Study of EFL Learners’ Critical Reasoning in Gemini-Assisted Genre-Based Writing
DOI:
https://doi.org/10.36312/jllet.v1i2.72Keywords:
EFL writing, Critical reasoning, Gemini, Generative artificial intelligence, Genre-based pedagogyAbstract
The increasing use of generative artificial intelligence in English as a Foreign Language (EFL) writing raises questions about whether learners critically evaluate AI-generated information or accept linguistically fluent outputs without sufficient scrutiny. This qualitative descriptive case study investigated how EFL learners critically evaluated information and arguments generated by Gemini during genre-based academic writing and how genre-based learning supported such reasoning. Seventeen EFL learners from private-school educational settings in Iraq were purposively selected based on their experience with online learning and academic writing. Data were generated through semi-structured interviews and classroom observations and analyzed iteratively through data condensation, data display, and conclusion drawing and verification. The findings revealed four interconnected dimensions of critical evaluation: assessing the credibility and accuracy of Gemini-generated information, evaluating evidence and argument strength, identifying assumptions and overgeneralizations, and judging genre appropriateness when deciding whether content should be accepted, rejected, or revised. Genre-based learning further supported these practices through building knowledge of the field, analyzing model texts, joint construction, independent text production, and reflection and revision. These stages provided conceptual, rhetorical, collaborative, and metacognitive scaffolding that shifted learners from passive acceptance toward selective and reasoned integration of AI-generated material. The study concludes that Gemini can contribute to academic writing when its outputs are treated as objects of critical inquiry rather than authoritative answers. Integrating explicit verification, evidence evaluation, rhetorical analysis, and reflective decision-making into genre-based pedagogy can help preserve learner agency and promote responsible human–AI collaboration in EFL academic writing.






