Students’ self-determination in using machine translation and generative AI tools for English for academic purposes

Abstract

The rise of machine translation (MT) and generative artificial intelligence (GAI) presents opportunities and challenges for English for Academic Purposes (EAP) instruction. While MT/GAI can support students' learning beyond the classroom, overreliance on MT/GAI may hinder development of essential research and composition skills. Despite some research on MT/GAI's role in self-regulated learning, little is known about students' motivations for its use, and the potential mediating influences of instructional context and discipline. This study applies Self-Determination Theory to examine how autonomy, competence, and relatedness influence students' use of MT/GAI in academic writing. Using a mixed-methods approach, the study compares EAP students in an English as a Second Language context in Australia and an English as a Foreign Language context in Vietnam. An online survey validated through confirmatory factor analysis and discriminant validity testing gathered 416 responses, complemented by interviews with 17 students. Findings reveal a complex interplay between MT/GAI use and students' motivational needs. While students generally report moderate autonomy, competence, and relatedness in using MT/GAI in the survey, mixed-effects regression showed Australian students experienced lower relatedness compared with Vietnamese students, with disciplinary differences also significantly influencing students' perceptions of this construct. Interview data further highlighted diversity and complexity of students' perceptions, variation in EAP instructional approaches and peer-teacher dynamics surrounding MT/GAI. These findings support the need for contextually tailored pedagogical approaches fostering collaboration between institutions, teachers, and students, illustrated through innovations from an Australian EAP course that bridge research and practice in MT/GAI-assisted academic writing.