The Effect of Generative Artificial Intelligence on Students’ Academic Achievement: A Meta-analysis

Generative artificial intelligence (GAI) technologyhas had a significant impact on teaching and learning, but its actual impact on students’ academic achievement is controversial. This study conducted a meta-analysis to systematically analyze 26 experimental and quasi-experimental articles on the impact of GAI on students’ academic achievement before April 2024. The results showed that: (1) GAI had a moderate positive effect on students’ academic achievement, (2) Compared with other disciplines, GAI can significantly improve the effect of medical disciplines, (3) The effect of GAI on students’ academic achievement was regulated by the sample size, and the medium sample had the higher effect value, (4) Multiple GAI applications over time are more effective than one-time intervention, (5) Compared to assessment feedback and combined application, conversational interactive GAI and content generative GAI are the most widely used and effective, and (6) GAI is applicable to various grade level and users.