EFL STUDENTS’ ATTITUDES AND PRACTICES TOWARD AI-BASED WRITING ASSISTANTS: AN EXPLANATORY MIXED-METHODS STUDY

The integration of artificial intelligence (AI) technologies in education has gained significant momentum, particularly through AI-based writing assistants (AIbWAs), which offer considerable potential for enhancing the writing skills of EFL students. This explanatory mixed-methods study aimed to investigate EFL students' attitudes toward AIbWAs and explore their practices in academic writing classes. The participants were 83 EFL students from a private university in Indonesia. Quantitative data were collected through a closed-ended questionnaire using a 4-point Likert scale to measure cognitive, affective, and behavioral dimensions of students’ attitudes. The AIWAs examined included QuillBot, Grammarly, and ChatGPT. Based on attitude scores, students were categorized into four groups: embracing (31.3%), acceptance (38.6%), indifference (26.5%), and resistance (3.6%). The mean attitude score was 3.21, reflecting a generally favorable perception. Students in the embracing and acceptance groups reported frequent use of AIbWAs for grammar correction, paraphrasing, vocabulary enhancement, and idea development. Students in the indifference group used the tools sporadically, often citing uncertainty or lack of confidence in the AI output. The resistance group raised concerns about overreliance, academic dishonesty, and the potential loss of personal writing voice. Twelve students representing each group were selected for semi-structured interviews. Thematic analysis of interview data revealed three key themes: perceived usefulness in improving writing quality, trust and skepticism toward AIbWAs' suggestions, and the perceived need for human judgment and ethical boundaries in academic writing. The findings underscore the importance of targeted digital literacy and pedagogical guidance to support effective and ethical use of AIbWAs in EFL contexts