English Majors’ Affective Attitudes Toward AI-Assisted Writing and Their Determinants

Abstract

Against the backdrop of rapid advancements in artificial intelligence technology, generative artificial intelligence (GenAI) has gradually been integrated into English writing instruction and plays a significant role in the writing feedback process; however, learners’ emotional experiences during its use remain to be explored in depth. This study focuses on undergraduate students majoring in English. Based on the ABC Attitude Model and employing semi-structured interviews and thematic analysis, it explores their emotional attitudes toward GenAI writing feedback and the factors underlying these attitudes. The results indicate that English majors generally exhibit a dual characteristic of positivity and caution toward GenAI writing feedback, which can be categorized into four types: positive affirmation, negative skepticism, conflicted ambivalence, and rational caution, with the rational caution type being the most central. While acknowledging its advantages, students also maintain a degree of vigilance regarding its shortcomings. The primary factors influencing the formation of their emotional attitudes include technology and feedback quality, individual differences, as well as disciplinary and situational requirements, with feedback quality being the key influencing variable. Furthermore, students’ emotional attitudes exhibit distinct situational and dynamic characteristics, adjusting according to different task types and writing stages. The study indicates that while GenAI optimizes writing support, it also elicits complex emotional experiences characterized by a coexistence of trust and dependence.