Developing the Anchor Eyes Model: An AI-Assisted Pedagogical Framework for Creative Poetry Writing in Secondary Education

Abstract

The rapid advancement of artificial intelligence (AI) has created new possibilities for writing instruction; however, its pedagogical integration into creative genres such as poetry remains underexplored. This study employed a research and development (R&D) design guided by the ADDIE instructional model to develop and examine the feasibility of the Anchor Eyes Model, an AI-assisted pedagogical framework for creative poetry writing in secondary education. The model was implemented with 96 secondary school students from three schools in East Java, Indonesia, within formal Indonesian language classes. Data were collected through performance-based poetry writing tasks, analytic scoring rubrics, student questionnaires, classroom observations, and semi-structured interviews. Quantitative data were analyzed using descriptive statistics to examine performance gains, while qualitative data were analyzed through thematic content analysis to capture students’ creative processes and perceptions of AI-assisted learning. The findings indicate meaningful descriptive improvements in students’ poetry writing performance, particularly in idea development, diction, and imagery, alongside positive learner perceptions of AI as a supportive creative scaffold rather than a substitute for original thinking. Qualitative evidence further revealed enhanced engagement, confidence, and reflective awareness during the creative writing process. These findings suggest that the Anchor Eyes Model offers a pedagogically grounded and ethically oriented approach to integrating AI into creative writing instruction. The study provides initial empirical evidence to inform human-centered AI pedagogy and supports the inclusion of structured AI-assisted models in secondary literacy curricula.