Is Generative Artificial Intelligence Effective in Improving Student Outcomes? A Meta-Analysis Based on 39 Experimental and Quasi-Experimental Studies

Abstract

The emerging generative artificial intelligence (GenAI) technology has the potential to become strong technological support for further reform and innovation in education by enriching instructional devices and modalities. Nevertheless, its specific effects on student outcomes remains contentious in academia. This article, using meta-analytical techniques, systematically reviews 39 experimental and quasi-experimental studies (encompassing 60 effect sizes) published before June 2024, aiming to investigate the effect of GenAI use on student outcomes and its moderating factors.