ChatGPT for L2 learning: Current status and implications

Abstract

Despite a recent proliferation of studies on ChatGPT for second language (L2) learning, there is still a lack of systematic and updated review of its current status. To narrow the gap, this study collected data from 44 selected studies on ChatGPT for L2 learning in terms of six dimensions of the revised technology-based learning model, including ChatGPT, participants, objectives, theories, methodology, and outcomes. The results showed that (1) The most prevalent ChatGPT's roles included content generation, feedback and teaching support. Context control and output customization were the two main prompt patterns. (2) Most studies focused on investigating English-as-a-foreign-language (EFL) college learners with small sample sizes. (3) Learner perceptions (general attitudes, satisfaction, motivation, and engagement) along with writing skills were the major objectives. (4) Social (sociocultural and constructivism), linguistic (input hypothesis, informal digital learning), and cognitive (self-determination theory, autonomy) theories were frequently adopted. (5) Most studies used qualitative, quantitative and mixed methods, with a particular eye on questionnaire surveys, interviews, log data and written texts. (6) Benefits and challenges were summarized from the selected studies. Implications were discussed for future research.