Learner agency in self-directed use of GenAI for informal language learning

Abstract

Informal learning contributes significantly to language development, and generative artificial intelligence (GenAI) may empower self-directed learning in informal contexts. Yet, agency is at the core of informal learning. As GenAI brings both affordances and constraints for learner agency, understanding how learners assert agency to partner with GenAI for self-directed informal learning is a significant area for exploration. This study adopted an exploratory sequential design to first unravel the dimensions of learner agency through thematic analysis of interview responses from 17 Chinese undergraduate English majors, and then validate the dimensions through confirmatory factor analysis and Rasch analysis on survey responses from 646 undergraduate English language learners. The study yielded a taxonomy of learner agency with GenAI for self-directed informal language learning, comprising five distinct dimensions–learning proxy agency, performance proxy agency, spatial agency, collective agency, and shared agency. These dimensions emerged in learners’ negotiation of subjectivity with socio-educational and sociotechnical structures within and around GenAI. This taxonomy provides a research and pedagogical tool to understand and support learner agency with GenAI for self-directed informal language learning.