Peer-reviewed article
A Meta-Analysis on the Effects of Coding Education for Early Childhood Education
Sunhee Kim ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This meta-analysis synthesizes research on coding education in early childhood settings, examining 26 studies published in Korean journals through 2024. The overall effect of coding education was large and positive, with the strongest effects on cognitive skills compared to physical and emotional development. Different coding approaches, including unplugged coding and coding robots, all showed positive effects, though differences between types were not statistically significant.
Key points
- Overall mean effect size of coding education for early childhood was 1.62, indicating large positive effects
- Pre-post experimental-control group designs showed larger effect sizes than comparative studies based on coding experience
- Cognitive domain showed the largest effect, followed by psychomotor and affective domains, with cognitive effects statistically significant
- Unplugged coding showed the largest effect size, followed by coding robots and integrated programs
- The study notes potential limitations from publication bias and small sample sizes in subcategories
Limitations
The meta-analysis only included studies published in Korean domestic journals, which limits generalizability to other contexts and languages. The authors acknowledge potential publication bias and small sample sizes in subcategory analyses, which may affect the reliability of comparisons between intervention types.
Abstract
While the importance of coding education in early childhood education has recentlybeen emphasized, research comprehensively examining its effects remains limited. The purpose of this study is to systematically analyze the effects of coding education for early childhoodeducation through a meta-analysis. The study subjects are research articles published inKorean domestic journals (KCI listed) up to 2024. The research method used is ameta-analysis following PRISMA guidelines, and effect sizes (Hedges' g) were calculatedusing EXCEL and R programs. The results are as follows: First, the overall mean effect size(g) of coding education for early childhood education was 1.62, with pre-postexperimental-control group designs showing larger effect sizes than comparative studies basedon coding experience. Second, regarding effect sizes across cognitive, affective, andpsychomotor domains, the cognitive domain showed the largest effect, followed bypsychomotor and affective domains, with the cognitive domain showing statistical significance. Third, regarding program types, unplugged coding showed the largest effect size, followed bycoding robots and integrated programs, although differences between types were notstatistically significant. This study empirically verified that coding education for earlychildhood education demonstrates large positive effects overall. The study discussed the needfor careful interpretation considering potential publication bias and small sample sizes insubcategories, and suggested implications and future research directions for effectiveimplementation of coding education for early childhood education.