<![CDATA[While traditional civic education often fails to bridge conceptual knowledge with active democratic behaviors, integrating Artificial Intelligence (AI) offers novel pathways for metacognitive scaffolding. This study evaluates the effectiveness of AI-scaffolded participatory reflection in nurturing primary school students' civic literacy and responsible character. Utilizing a quasi-experimental pretest-posttest control group design, the research involved fifth-grade students (N = 64) divided into an experimental group (n = 32) employing a generative AI chatbot to scaffold reflection and a control group (n = 32) undergoing conventional learning. Data from validated tests, questionnaires, observations, and interviews were analyzed via ANCOVA to control initial variances. Inferential results demonstrated that the AI-scaffolded approach significantly outperformed conventional instruction in enhancing civic literacy (F(1, 61) = 39.56, p < 0.001, d = 1.21) and developing responsible character (F(1, 61) = 55.06, p < 0.001, d = 1.34). The experimental group achieved higher normalized gains, accelerating task completion and behavioral self-regulation. Qualitative insights confirmed that real-time AI feedback facilitated autonomous error correction while mitigating automation bias. Theoretically, this study contributes an empirical model merging AI-mediated metacognition with dialogic pedagogy, transitioning citizenship education from passive memorization to active character internalization. Pragmatically, it provides a scalable framework supporting United Nations’ Sustainable Development Goal (SDG) 4.]]>
Copyrights © 2026
