Gender disparities in computational thinking (CT) education are widely acknowledged, but few meta-analyses have investigated how particular instructional approaches and assessment settings shape these differences. To address this research gap, Liu et al. (2025) conducted a meta-analysis of 53 empirical studies, covering 100 effect sizes and a total sample of 15,454 participants, to examine the overall magnitude of gender differences in CT education and the factors that may shape them. The findings show a small but statistically significant overall gender difference (g = 0.106, 95% CI [0.024, 0.188], p < .05), suggesting a slight advantage for males.
Regarding moderation effects, neither general study features (e.g., publication type, geographic region, and educational level) nor CT assessment contexts (e.g., the instrument used and the learning outcome measured) significantly altered the effect sizes. In contrast, pedagogical approaches did matter: technology-integrated strategies such as mixed and plugged approaches were linked to larger gender gaps favoring boys, whereas unplugged approaches tended to narrow the gap and sometimes even shifted the advantage toward girls. In terms of assessment, gender differences were not significant when CT concepts were measured, but they became significant when outcomes involved authentic practices (such as programming tasks) and identity-related dimensions (such as motivation, learning interest, and self-efficacy).
The results highlight clear implications for improving equity in CT education. Support should start early in K–12, with a particular focus on developing students’ CT practices and perspectives so that small gender gaps do not become persistent over time. Unplugged activities can serve as a low-barrier entry point, strengthening basic understanding and confidence, especially for girls. In addition, technology use should be introduced progressively: when digital and AI tools are scaffolded within supportive, culturally relevant learning settings, students may feel less anxious about technology and experience more inclusive participation.
Source (Open Access): Liu, S., Dai, Y., Ng, O. L., & Cai, Z. (2025). Gender Disparity in Computational Thinking Pedagogy and Assessment: A Three-Level Meta-Analysis. Educational Psychology Review, 37(4), 114.