Abstract

This meta-analysis examined the effects of sex education programs for children and adolescents conducted in Korea between 2000 and 2024 on perceptions of sex, and analyzed the structure of these effects. From 33 journal articles and theses, 76 effect sizes were extracted by dependent variable. Perceptions of sex were reclassified into four subdomains-sexual knowledge, sexual attitudes, sexual awareness, and self-understanding-based on the content of the measurement items. Hedges' g, adjusted for small-sample bias, was computed under a random-effects model. The overall mean effect size was 1.07 and statistically significant. Large effects were observed for sexual knowledge, sexual attitudes, and sexual awareness, while the effect on self-understanding was comparatively small, and between-subdomain differences were significant. In contrast, subgroup analyses by number of sessions and age group yielded no significant moderating effects. These results suggest that program effectiveness depends more on qualitative factors—program content, instructional methods, instructor expertise, and measurement characteristics than on quantitative conditions. By applying item-content-based coding to address the conceptual ambiguity and measurement inconsistency long noted in this literature, this study systematically delineates the multidimensional structure of sex education effects on perceptions of sex.