Exploring The Use of Text-Generative AI in Supporting Self- Regulated Learning in Undergraduate Thesis Writing in English Education Study Program

Abstract

ABSTRACT

Researcher : Rahmat Reg. Number : 10256121034

Title : Exploring The Use of Text-Generative AI in Supporting Self- Regulated Learning in Undergraduate Thesis Writing in English Education

Study Program The increasing integration of artificial intelligence (AI) in higher education has transformed how students plan, write, and revise academic texts. While text-generative AI offers various affordances, its influence on students’ ability to regulate their own learning during complex writing tasks such as thesis writing remains underexplored. This study

explores how undergraduate students in the English Education Study Program utilize text- generative AI tools to support their Self-Regulated Learning (SRL) during thesis writing.

Using a qualitative descriptive method, data were collected from seven students through semi-structured interviews and analyzed with NVivo 12 Plus. The analysis identified three main phases of SRL based on Barry J. Zimmerman’s model, forethought, performance, and self-reflection, each consisting of specific themes and strategies related to AI use. In the forethought phase, students used AI to analyze tasks, set goals, develop thesis outlines, and build confidence before writing. During the performance phase, they applied AI tools to draft and revise their work, generate ideas, structure sentences, refine language, evaluate content, and implement targeted prompting strategies. In the self-reflection phase, students relied on AI to evaluate drafts, incorporate supervisors’ feedback, and check final coherence, while also becoming aware of the risks of over-reliance on AI. These findings reveal that students did not merely copy AI outputs but integrated AI as part of their internal learning regulation system to plan, monitor, and evaluate their writing processes. This study contributes to the understanding of how text-generative AI can serve not only as a technical writing aid but also as a motivational and metacognitive scaffold that enhances students’ autonomy and self-regulation. However, it also highlights the need for balanced AI use to prevent dependency and maintain critical engagement with content. The findings offer practical implications for guiding students in adopting AI tools effectively and ethically in academic writing contexts. Keywords: Writing Process, Educational Technology, Zimmerman’s SRL Model, ChatGPT, Grammarly.