Whether English Proficiency and English Self-Efficacy Influence the Credibility of ChatGPT-Generated English Content of EMI Students

Abstract

This study investigated whether English proficiency (EP) and English self-efficacy (ESE) influence the credibility of ChatGPT-generated English content (CCGEC) of college students in the English-medium instruction (EMI) courses. We observed a college-level EMI course called “Technical Writing and Presentation”, where instructors encouraged their students of non-English majors to use ChatGPT for discovery learning in their EMI learning tasks. A survey of 800 students showed that ChatGPT was primarily used for providing explanations, suggesting ideas and structures, and preparing presentations. Analyzing the survey data with the MANCOVA procedure revealed that students' EP levels did not influence their CCGEC, and ESE did not moderate the relationship between EP and CCGEC in the EMI course context. These findings suggest that ChatGPT can be a valuable tool for EMI students regardless of their EP or ESE.