top of page

Volume 3· Number 4 · 2026

Teaching Evaluation and Measurement

AI-Enhanced Formative Assessment Feedback Loop for Primary Mathematics: Development and Validation of the AI-FAF Model in a Korean Elementary School

Park Ji-hoon

Gangnam Elementary School, Gangnam-gu, Seoul, Republic of  Korea

 

 

Abstract: This study presents the development and validation of an AI-Enhanced Formative Assessment Feedback Loop (AI-FAF) model designed to improve formative assessment practices in primary mathematics education. Developed and implemented by a classroom teacher at Gangnam Elementary School in Seoul, Korea, the AI-FAF model addresses a critical limitation of traditional formative assessment—its focus on final answers rather than mathematical thinking processes—by leveraging artificial intelligence to analyze students' problem-solving processes and generate adaptive, process-focused feedback. The model features four interconnected components: (1) an AI Process Analyzer that examines solution strategies, error patterns, and reasoning pathways; (2) an Adaptive Feedback Generator that provides multi-layered feedback matched to student needs; (3) a Teacher Dashboard that translates AI analytics into actionable instructional insights; and (4) a Student Self-Assessment Portfolio that supports metacognitive development. The study employed a mixed-methods design with 87 Grade 5 students across three classes, implementing the model over a 12-week period focusing on multiplication and division of fractions. Quantitative results indicate that students in the experimental group showed significantly greater improvement in both procedural fluency (t=4.82, p<.001, d=0.73) and conceptual understanding (t=5.16, p<.001, d=0.78) compared to the control group. Qualitative data from classroom observations, teacher journals, and student interviews further reveal that the AI-FAF model enhanced the quality of feedback, supported more targeted instruction, and promoted student metacognition. The study contributes to the growing Korean literature on AI-enhanced formative assessment and provides a practical, teacher-developed model that can be adapted to diverse primary mathematics contexts.

 

Keywords: formative assessment, AI in education, feedback, primary mathematics, process-focused assessment, Korean education, self-assessment


ISSN: 3066-229X 版权所有 © 2024  Reviews Of Teaching

bottom of page