Fostering learner autonomy through data-driven learning in middle school students’ acquisition of English collocations

Data-driven learning (DDL), which utilizes authentic language samples from corpora, promotes independent learning by allowing students to discover language patterns. While previous studies highlight DDL’s effectiveness in language acquisition, its role in fostering learner autonomy remains underexplored. This study investigates the impact of DDL on autonomy and the acquisition of verb-preposition collocations among low-proficiency Korean middle school students. Over twelve weeks, five students engaged in hands-off (teacher-guided) and hands-on (independent) DDL activities, analyzing collocations using two corpus tools. Data were collected through pre- and post- tests, classroom observations, reflective journals, screenshots, and interviews. Findings suggest that DDL improved learners’ ability to construct verb-preposition collocations and fostered greater autonomy through self-directed problem-solving, increased learner agency and participation, and strategic engagement with corpus tools. While learners initially faced challenges in navigating corpus tools, they gradually developed effective learning strategies and greater confidence in independent exploration. These results underscore DDL’s potential to enhance both language competence and autonomous learning, particularly for low-proficiency learners.