Meta-analysis on effects of artificial intelligence education in K-12 South Korean classrooms

Meta-analysis on effects of artificial intelligence education in K-12 South Korean classrooms

연구 분야: Verification

학회: Education and Information Technologies

This study aimed to integrate the results of prior studies on the effectiveness of AI education in K-12 Korean classrooms to draw systematic and comprehensive conclusions. To achieve this goal, a review of 64 studies on AI education in Korea that were conducted from 2019 to 2023 was subjected to a meta-analysis. The total effect size of AI education and the effect size for each categorical variable were calculated. The results of the meta-analysis are as follows. The overall effect size of AI education is 0.897 (95%CI = 0.781 ~ 1.013, p < .05), which can be interpreted as a large effect. In terms of learning effect, the cognitive domain had a large effect, and the affective domain had a medium effect. In addition, the effect size by publication type, school grade, subject, class type, activity type, and other categories was presented. In the early stages of AI education policy, this meta-analysis presents evidence-based action recommendations by synthesizing the results of prior studies. This study provides implications for instructional design and supportive plans that can increase the effectiveness of AI education.