Abstract
The aim of this study is to systematically evaluate, through a meta-analysis, the impact of STEM instruction based on solar energy on students’ research abilities. In the study, 10 empirical works published over the past decade were selected, and their standardized effect sizes were synthesized. The analysis was conducted using a random-effects model. The results showed a pooled mean effect size of d = 1.488 (95% CI: 0.942–2.035, p < 0.001), indicating a strong positive effect of STEM interventions in the context of solar energy on research abilities. Heterogeneity among studies was at a moderate level (I² = 58%), suggesting that the magnitude of the effect may vary depending on the educational context. The results of publication bias assessment confirmed the stability of the overall effect. In addition, meta-regression analysis showed that the most pronounced effect of the STEM approach was observed in the domain of research skills (p < 0.01). To complement the meta-analytic findings, an additional cross-sectional correlational study was conducted with 49 students who participated in a solar energy-based STEM learning program. Students’ engagement in STEM activities and their research competence were assessed using two composite indicators, STEM Level and Research Activity, each standardized on a 0–100 scale. Pearson’s correlation analysis was used to examine the relationship between these indicators. The complementary empirical analysis revealed a strong positive correlation between STEM Level and Research Activity (r = 0.807, p < 0.001), providing convergent contextual evidence consistent with the direction of the meta-analytic findings.