Effects of Quillbot-assisted paraphrasing practice on EFL college students’ English vocabulary acquisition and perception

Objectives This study explored the effects of Quillbot-assisted paraphrasing practice on EFL college students’ English vocabulary acquisition and their attitude toward of Quillbot-paraphraser use. Methods Divided into two, the experimental group (n = 17) performed paraphrasing writing tasks with Quillbot-paraphraser whereas the control group (n = 18) completed the same tasks without it. The two groups took vocabulary quizzes at two-week intervals and the experimental group responded to the questionnaire concerning their perception toward Quillbot-paraphraser use. By Mann-Whitney U test, comparisons were made on each quiz score between the groups and the responses on the questionnaire of the experimental group were analyzed by content. Results The result found no statistically significant differences in the vocabulary quiz scores of the two groups, but the analyses of the students of the experimental group showed that Quillbot-paraphraser use certainly helped their paraphrasing performance, extending their vocabulary knowledge in depth, aiding them to comprehend sen-tence structures better, and making them aware of distorted sentences. Conclusions There remained limitations in that the Quillbot-paraphraser provided distorted sentences that were different from the students’ intended meanings and was unable to provide detailed explanations related to its vo-cabulary choices. Thus, it seems important students themselves to have an keen eyes to identify problems with the results provided by AI tools. It is hoped that further studies will lead to in-depth discussions on the new role of teachers and on discoveries of new strategies for implementing AI tool-assisted writing tasks, calling teachers’ facilitating supports in the classrooms.