Corpus-informed vocabulary instruction in Bangladesh: a practical framework for EFL contexts

Abstract This study details the design and classroom implementation of a 36-week, corpus-informed vocabulary programme developed for Bangladeshi EFL learners. The study aimed to create a practical, replicable teaching framework to enhance academic vocabulary acquisition in resource-limited educational settings. The intervention combined corpus consultation tasks, guided collocation discovery, and production-focused writing activities integrated into the standard English curriculum. A single-group pre–post feasibility study was used to assess the practicality of classroom delivery, teacher workload, and observable language improvements. Participants completed writing tasks before and after the intervention, which were analysed for academic vocabulary coverage, collocation accuracy, and use of formulaic sequences. Descriptive analysis showed notable lexical and collocational gains, supported by teacher reflections and student focus groups. Rater agreement for coding was high ( κ = 0.82), indicating reliability. Results suggest that corpus-informed vocabulary instruction can be feasibly incorporated into existing curricula when materials and teacher support are adapted to local contexts. Although causal conclusions cannot be drawn without a control group, this study offers an evidence-based procedural model and transparent analytical framework for future large-scale, controlled research in EFL settings.