POTENTIAL RISKS OF CHATGPT-ASSISTED ESSAY WRITING ON KNOWLEDGE RETENTION AMONG EFL LEARNERS IN INDONESIA

Abstract

This study investigates the impact of ChatGPT-assisted essay writing on knowledge retention among English as a Foreign Language (EFL) learners, comparing it with traditional Google-based research and a hybrid approach that combines both tools. A total of 150 undergraduate students from an Indonesian university participated in the study and were divided into three groups: ChatGPT-only, Google-only, and hybrid. To measure knowledge retention, a 20-item multiple-choice post-test was administered one week after the essay-writing activity. For the qualitative phase, open-ended online surveys were used to explore students’ perceptions of the tools, learning experiences, and challenges encountered. Using a mixed-methods approach, quantitative data were analyzed through descriptive statistics and ANOVA, revealing that the Google-only group achieved the highest mean score, followed by the hybrid group, while the ChatGPT-only group had the lowest performance. Thematic analysis of qualitative data identified four key themes: efficiency, depth of understanding, challenges, and learning outcomes. Results suggest that when ChatGPT supports efficiency in writing, it often limits critical engagement and deep learning. In contrast, Google encourages active information processing, enhancing content comprehension and long-term retention. The hybrid approach offers a balanced strategy by combining the strengths of both tools. These findings contribute to the growing discourse on AI in education by emphasizing the need to evaluate not only usability and writing quality but also cognitive outcomes such as retention and understanding. The study highlights the pedagogical importance of integrating AI tools thoughtfully, rather than relying on them as stand-alone solutions for academic tasks