Harnessing AI-based learning media in education: A meta-analysis of its effects on student achievement

From the most straightforward kind of technology—audiovisual learning—to the application of artificial intelligence in education, technology has been used in education for over 20 years. Despite the growing popularity of AI-based learning media technology, there is still a dearth of reliable empirical data about its effects on student achievements. This meta-analysis aims to investigate the impact of intervention time and combine findings from several studies to paint a more comprehensive picture of the usefulness of AI media in education. In this study, a meta-analysis design is employed in quantitative research. The Publish or Perish tool gathered secondary data from published papers using the Scopus database and Google Scholar—data analysis for group contrast meta-analysis data using the R software. The study's findings demonstrate how using AI-based learning resources greatly impacts students' academic performance. P value total effect size and three moderator variables (continent, gained achievement, and intervention duration < 0.05) show that the aggregate value of the summary effect in AI-based learning media, which integrates technology products with software, web programs, augmented reality, and gamification in increasing student achievement from elementary school to tertiary level from 2019 to 2024, is still providing significant influence. Thus, artificial intelligence (AI) should be used more extensively in preparing learning media to maximize students' academic and non-academic successes.