Toward “AI-Centered Student”: Making Sense of Learning Environment, Epistemic Beliefs, and Self-Regulated Learning Using Epistemic Network Analysis

Abstract

This preliminary study is designed to explore the multifaceted impact of Generative Artificial Intelligence (GenAI) integrated into digital learning platforms, namely AISI, on students' science learning, self-regulation, and epistemic beliefs from 129 college students. Leveraging empirical study data, Epistemic Network Analysis (ENA) was utilized to quantitatively explore relationships between students' learning preferences (high AI-centered preference group and low AI-centered preference group), their epistemic beliefs concerning AI (certainty, justification, complexity), and their self-regulated learning strategies (adaptation, planning). The findings show that students in the high AI-centered preference group tend to believe that knowledge provided by GenAI is uncertain and complex. This study aims to contribute to GenAI-enhanced learning environments and pedagogical practices that foster critical AI literacy and adaptive self-regulated learning.