E-Portfolio Development Literacy in the GenAI Era: A Comparative Mixed-Methods Study

The increasing integration of generative Artificial Intelligence (GenAI) into higher education has reshaped language assessment, drawing growing scholarly attention to process-oriented approaches that can capture learning development while accommodating AI-mediated writing practices. E-portfolio assessment has gained significant momentum in this context. Yet, research has rarely examined how different e-portfolio media shape the ways students create, curate, review, and disseminate multimodal artefacts. Drawing on the framework of E-portfolio development literacy (Lam, 2025), which encompasses the knowledge, skills, and dispositions required for effective e-portfolio compilation, this proposed study investigates how the choice of e-portfolio medium influences EFL students’ literacy development in a GenAI-mediated formative assessment environment. Adopting a quasi-experimental explanatory mixed-methods design, the study will compare three e-portfolio media: a learning management system (Xuexitong), a social media platform (RedNote), and a GenAI-integrated workspace (Notion AI). Three classes of first-year non-English-major students at an Eastern Chinese university will participate, with GenAI incorporated into all conditions. Over a 16-week semester, students will compile e-portfolios to document their learning trajectories through written assignments, video productions, and group showcases. Quantitative data, to be collected through pre- and post-intervention measures of language proficiency, feedback literacy, and critical AI literacy, will be analysed using paired-sample t-tests and ANCOVA. Qualitative data, including reflective writing, interviews, narrative inquiry data, AI chatlogs, and multimodal artefacts, will undergo thematic, content, and multimodal analyses, with cross-medium comparisons and triangulation across data sources. This study positions the e-portfolio medium as an active driving force that documents and accommodates compilation processes rather than merely a delivery platform. It offers both a theoretical and empirical foundation for advancing the scholarship of e-portfolio-based formative assessment, which, in turn, validates and supports EFL students’ language learning in GenAI-mediated contexts.