Developing Pre‐Service Teachers’ Digital Self‐Efficacy Through Self‐Directed Language Learning

ABSTRACT Contemporary online language learning typically involves the use of multiple digital platforms. Drawing on digital self‐efficacy and self‐directed learning theories, this study investigates the self‐directed language learning experience of 19 pre‐service teachers as they navigate the affordances of two platforms, Duolingo and ChatGPT. We collected data through participants’ theory‐informed reflections on their learning experiences and screenshots of key learning moments over a 12‐week semester. Participants demonstrated their ability to navigate between aspects considered to promote or impede learning on each platform. Their learning experiences were influenced by extrinsic factors of the learning environment (e.g., distractions) and the learning platforms (e.g., the structured design of Duolingo, the instant feedback of ChatGPT). Participants employed metacognitive strategies to adapt to their learning context and needs, to maximise the affordances of each platform for different learning purposes (Duolingo for structured learning and ChatGPT for flexible and personalised learning), and to monitor their learning outcomes. Participants documented the refinement of their prompt formulation and sequencing abilities. Time management was closely linked to the platforms’ affordances, participants' personal motivation and their susceptibility to distractions. Participants displayed lower digital self‐efficacy when relying on a single platform, but they gained confidence when using both platforms for complementary learning. We close with implications for enhancing digital self‐directed language learning.