Speaker
Description
Generative artificial intelligence (GenAI) is increasingly being integrated into second language (L2) writing instruction, yet teachers remain the key decision-makers in determining whether AI-generated feedback is accurate, pedagogically appropriate, and suitable for classroom use. While prior research has focused primarily on the quality of AI feedback or students’ uptake, less is known about how different prompting strategies shape teachers’ evaluations and which teacher characteristics influence their trust and intended use of AI-generated feedback. This study examines how pre-service and in-service English teachers evaluate AI-generated feedback produced under four prompting conditions: zero-shot, few-shot, structured stepwise, and teacher-scaffolded prompting. Using a sequential explanatory mixed-methods design, 40 teacher participants evaluated AI-generated feedback on argumentative essays on a five-point Likert scale and completed a questionnaire measuring GenAI literacy, prompting confidence, prior AI experience, perceived feedback quality, trust, and intended instructional use. Follow-up semi-structured interviews further explored teachers’ evaluative reasoning. Quantitative data were analyzed using mixed-effects models, while qualitative data underwent thematic analysis to explain and enrich the quantitative findings. The preliminary findings showed that structured stepwise and teacher-scaffolded prompting received more favorable evaluations than zero-shot and few-shot prompting, particularly in terms of pedagogical usefulness, clarity, and trust. Higher levels of GenAI literacy and prompting confidence were most strongly associated with greater trust in AI-generated feedback and stronger intentions to integrate it into classroom practice. The findings are expected to inform teacher education programs and provide evidence-based guidance for designing effective prompting strategies that support the responsible integration of AI feedback into L2 writing instruction.
Keywords: AI-generated feedback, L2 writing, prompting strategies, teacher evaluation
