Habiburrahman Habiburrahman
Master of Islamic Education Management, Universitas Islam Negeri Sunan Kalijaga, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

University Students’ Perceptions of AI-Generated, Human, and Hybrid Feedback in Academic Writing Anta Alaka; Rahmatiya Husain; Tiya Dirmayani; Habiburrahman Habiburrahman; Eka Widayanti
JINoP (Jurnal Inovasi Pembelajaran) Vol. 12 No. 1 (2026): Mei
Publisher : University of Muhammadiyah Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22219/jinop.v12i1.42215

Abstract

Background: The increasing use of artificial intelligence (AI) tools in higher education has transformed the way students receive feedback on academic writing. While previous studies have compared AI-generated feedback and human feedback, limited attention has been given to students’ perceptions of the effectiveness of different feedback sources, the writing aspects influenced by each type of feedback, and preferences for hybrid feedback approaches. Objectives: This study aimed to investigate university students’ perceptions, preferences, and experiences regarding AI-generated feedback and human feedback in academic writing. Methods: This study employed a descriptive survey design involving 123 university students with experience using AI-assisted writing tools and receiving human feedback in academic writing. Data were collected through a structured questionnaire and analyzed using descriptive statistics, including frequencies, percentages, means, and standard deviations. Results: The findings revealed that both AI-generated feedback and human feedback were perceived as beneficial for academic writing development. However, human feedback received a higher effectiveness rating (M = 4.19, SD = 0.86) than AI-generated feedback (M = 3.65, SD = 0.92). AI-generated feedback was primarily associated with improvements in grammar, spelling, sentence structure, and language accuracy, whereas human feedback was perceived as more beneficial for argument development, coherence, and higher-order writing skills. The results also showed that hybrid feedback was the most preferred approach (43.1%), followed by AI-only feedback (36.6%) and human-only feedback (20.3%). Participants identified speed, accessibility, and consistency as the main advantages of AI-generated feedback, while limitations included difficulties in understanding academic context, interpreting nuanced arguments, and supporting critical thinking. Conclusions: The findings suggest that AI-generated feedback and human feedback serve complementary roles in academic writing development. While AI-generated feedback effectively supports lower-order writing concerns, human feedback remains essential for higher-order writing skills. Therefore, integrating AI-generated feedback with human feedback may provide a more balanced and effective approach to academic writing instruction in higher education.