ABSTRACT
An updated version of Google Translate (GT) incorporating a system of artificial intelligence, Google Neural Machine Translation (GNMT) was launched in 2016. Some studies have been conducted to examine the use of the upgraded GT as an educational tool, especially for college students majoring in English. The present study focused on implementing GT into academic English writing for Chinese non-English major university students to understand its effectiveness, advantages, and impact. Students successively composed an essay in Chinese, completed a corresponding composition in English, translated their Chinese text into English using GT, and finally revised their self-written (SW) versions by referring to their GT-translated versions. Three types of online computational assessments were used to evaluate their writing. Results indicated that the GT-translated versions were significantly better than the SW versions. Students significantly improved their English writing in the revised self-written (RSW) versions after interacting with instant feedback from their individual GT-translated versions, with an even greater improvement for the students with lower English proficiency. Students also held positive perceptions about the use of GT as a useful tool in academic EFL writing.
