CHATBOT AI SEBAGAI DIGITAL SCAFFOLDING UNTUK MENDUKUNG SELF-REGULATED LEARNING PADA PEMBELAJARAN INKUIRI DI SEKOLAH DASAR

Abstract

The low level of learning independence among elementary school students, as reflected in excessive dependence on teacher instructions, limited initiative in exploration, and weak self-monitoring skills during the inquiry process, has become a fundamental issue that has not been systematically addressed through appropriate educational technology approaches. This study aimed to examine the effect of using an AI chatbot as digital scaffolding on improving the learning independence of fifth-grade elementary school students in the subject of Mathematics through the inquiry learning model. The study employed a quasi-experimental research method with a Nonequivalent Control Group Design, involving 60 fifth-grade students from SDN 060914 Sunggal who were divided into two groups, consisting of 30 students in the experimental class and 30 students in the control class, selected through purposive sampling. The research instruments included a four-point Likert-scale learning independence questionnaire, an inquiry activity observation sheet, and chatbot interaction documentation. The results showed that the average learning independence score in the experimental class increased from 61.23 to 79.87 with an N-gain of 0.52 (moderate category), while the control class only achieved an N-gain of 0.24. The independent samples t-test produced a value of t = 6.14 with p < 0.001, and Cohen’s d effect size of 1.14, which was categorized as large. These findings indicate that AI chatbots positioned as digital scaffolding significantly facilitate the development of elementary students’ learning independence, particularly through adaptive prompting mechanisms, guiding questions, and immediate feedback that support self-regulated learning processes within the*inquiry learning framework.