“Why Isn’t Anything Showing Up?”: Interactional Practices and the Development of Corpus Literacy in Data-Driven Writing

Corpus-based data-driven learning is widely recognised in applied linguistics, yet how students interact with and around corpus tools in Vietnamese classrooms remains underexplored. Grounded in social constructivist learning theory, this study investigates how Vietnamese EFL students engage with a corpus tool during collaborative writing, and the difficulties and strategies emerging from this process. The study involved six English-major undergraduates working in small groups during a 60-minute corpus-based writing session in a Vietnamese university classroom. Students’ interactions with the corpus tool, peers, and the teacher were video-recorded and analysed qualitatively. The findings show that corpus use unfolded along two closely connected dimensions: operational interaction , related to navigating the corpus and retrieving data, and interpretive interaction , related to making sense of corpus evidence for writing decisions. Students also encountered technical difficulties during corpus use. These difficulties were managed through peer-supported step recovery, evidence-based comparison, and – when uncertainty persisted – teacher mediation. Rather than indicating failure, such difficulties reflected learners’ emerging corpus literacy. The data also reveal learners occasionally oriented to other digital tools such as Google search or AI-based writing assistants as potential resources for verifying corpus findings. Overall, the study highlights the interactional and developmental nature of corpus-assisted writing and emphasises the importance of structured support and flexible use of multiple tools during corpus-based instruction.