AI-Assisted Research Writing: Graduate Students' Experiences, Outcomes and Academic Integrity
This study examined the relationship between artificial intelligence (AI) tool use and writing outcomes among graduate students engaged in thesis writing, grounded in the Technology Acceptance Model. Using a convergent parallel mixed methods design, the study involved 96 graduate students enrolled in thesis writing courses at a government higher education institution in Bulacan, during the first trimester of the 2025–2026 academic year. Qualitative data focused on students' adherence to academic integrity regulations. Quantitative measures assessed perceived usefulness, perceived ease of use, behavioral intention to use, and self-reported writing performance. The results indicated that ChatGPT, Grammarly and QuillBot were the most used AI tools. Statistical analysis demonstrated a strong positive relationship between experience with AI tools and writing performance outcomes. The qualitative analysis surfaced themes about AI tool adoption, namely linguistic support, organization of ideas, and cognitive scaffolding and ideation¾alongside themes reflecting ethical awareness, namely selective use and resistance, responsible and transparent authorship, critical evaluation, accountability in AI tool use, and alignment with institutional guidelines. These results show that technological advances are now integral to writing courses and are used in the preparation and oversight of students' research. The study's key implication is that to use AI tools in thesis writing in an ethical manner, institutions must develop clear ethical frameworks that protect scholarly integrity while accepting the changing reality of human–AI collaboration in academic research.
Paper
Full text
AI-Assisted Research Writing: Graduate Students' Experiences, Outcomes and Academic Integrity
Semantic Scholar · 2026
Abstract
This study examined the relationship between artificial intelligence (AI) tool use and writing outcomes among graduate students engaged in thesis writing, grounded in the Technology Acceptance Model. Using a convergent parallel mixed methods design, the study involved 96 graduate students enrolled in thesis writing courses at a government higher education institution in Bulacan, during the first trimester of the 2025–2026 academic year. Qualitative data focused on students' adherence to academic integrity regulations. Quantitative measures assessed perceived usefulness, perceived ease of use, behavioral intention to use, and self-reported writing performance. The results indicated that ChatGPT, Grammarly and QuillBot were the most used AI tools. Statistical analysis demonstrated a strong positive relationship between experience with AI tools and writing performance outcomes. The qualitative analysis surfaced themes about AI tool adoption, namely linguistic support, organization of ideas, and cognitive scaffolding and ideation¾alongside themes reflecting ethical awareness, namely selective use and resistance, responsible and transparent authorship, critical evaluation, accountability in AI tool use, and alignment with institutional guidelines. These results show that technological advances are now integral to writing courses and are used in the preparation and oversight of students' research. The study's key implication is that to use AI tools in thesis writing in an ethical manner, institutions must develop clear ethical frameworks that protect scholarly integrity while accepting the changing reality of human–AI collaboration in academic research.