Toward Agent-Based Educational Science: Rethinking Educational Research in the Age of AI

Abstract

Educational science faces a structural mismatch between the pace of educational innovation and the methods used to evaluate its developmental impact. While new pedagogical approaches and AI-driven learning technologies are rapidly deployed, the empirical paradigm of classroom-based research remains slow, fragmented, and ethically constrained, often generating evidence only after large-scale implementation has occurred. Here, we argue that education requires a paradigmatic shift toward agent-based educational science: a research framework in which educational theories are formalized as interacting agents and environments, enabling in silico experimentation on developmental processes that are otherwise slow, risky, or infeasible to test empirically. Recent advances in generative artificial intelligence make such a shift practically achievable for the first time. As a concrete instantiation of this paradigm, we introduce Student Development Agents—computational agents designed to generate longitudinal developmental trajectories under counterfactual educational environments. Rather than replacing empirical research, agent-based educational science reconfigures its role, enabling predictive, ethical, and cumulative theory building in the science of learning.