Journal Paper Published

Abstract

Providing effective writing feedback to students could promote students' writing development. However, offering high-quality feedback remains a significant challenge for pre-service teachers (PSTs). Recent advancements in GenAI technology may offer solutions to this issue. The study examines the influence of a short-term teaching intervention on the feedback levels and feedback types of PSTs' ChatGPT-assisted feedback on students' written compositions, utilizing an explanatory sequential mixed-method design with 30 PSTs. The quantitative results revealed significant improvements in both feedback levels (i.e., higher-level feedback issues including ideas and elaboration and style) and feedback types (e.g., explanations and general suggestions). Additionally, the findings highlighted specific strategies employed by PSTs when considering levels and types in combination. Subsequent interviews identified the underlying influential factors of these improvements, namely the improvements in ChatGPT usage skills (i.e., prompt engineering and source use) and a deeper understanding of the feedback process (i.e., introspection). By demonstrating how short-term teaching interventions can leverage PST's ability to use GenAI tools to provide writing feedback, this research advances the theoretical understanding of human-AI collaboration in the context of writing and provides pedagogical insights for teacher training programs.

Keywords