Yunan Zhang, Yongcan Liu
With the rise of GenAI-assisted English as a Foreign Language (EFL) writing, chatbots such as ChatGPT have shown potential both as formative feedback providers and interactive partners, offering new possibilities for dialogic feedback. However, little research has explored how educational practitioners can design and implement ChatGPT-mediated dialogic feedback activities. Addressing this gap, this study adopted a design-based research (DBR) approach to iteratively develop an optimal design for ChatGPT-mediated dialogic feedback activities, drawing on a dialogic feedback triangle as a conceptual framework. Across three iterative cycles with four EFL learners, data were collected through semi-structured observations and interviews focused on participants’ writing experiences and feedback on the intervention. The three research rounds led to the identification of seven key design elements, organised into a three-dimensional feedback triangle: cognitive (disciplinary knowledge, cue-consciousness, self-evaluation), socio-affective (adaptive roles of ChatGPT at different writing stages, learners’ critical attitudes towards ChatGPT), and structural (mobilisation of learners’ tool repertoires, feedback structures based on specific formats and standards). These design elements can offer insights into feedback literacy and inform seven practical deisign principles to guide educators and technology designers in designing and implementing effective GenAI-mediated dialogic feedback experiences in EFL writing contexts.