Enhancing Higher Education through AI-Driven Personalized Adaptive Learning: Evidence from a ChatGPT-Based Strategy

Abstract

Artificial intelligence (AI) is reshaping higher education through personalized adaptive learning systems that tailor instruction to students’ needs. Most evidence focuses on short-term outcomes in STEM; social sciences remain understudied. This study evaluates a ChatGPT-based personalized adaptive learning strategy in an undergraduate Political Science course at a Chinese university. The 16-week intervention integrated AI-driven activities into regular coursework. Outcomes were measured with structured assessments and student surveys. Using a quasi-experimental design with propensity score matching and difference-in-differences, we examined learning efficiency, knowledge mastery, and student satisfaction. The experimental group significantly outperformed the control group across all dimensions. Observed gains were associated with adaptive learning paths, timely feedback, and interactive engagement. These findings suggest that ChatGPT may offer benefits for social science education and may inform the design of AI-powered personalized learning in higher education. The study extends the evidence base on AI beyond STEM and highlights the importance of rigorous evaluation in real course settings.