Due to the interdisciplinary nature of robotics, more and more attention has been paid to its effectiveness in the field of education in recent years. This systematic review evaluated existing studies in improving K-12 students’ computational thinking and STEM attitudes. Research articles published between 2010 and 2019 were collated from major databases according to six criteria, and 17 studies were eligible. A meta-analysis was conducted to evaluate the effectiveness of educational robots in terms of standardized mean differences (SMD) or mean differences (MD) of test scores as outcome measures. The overall effect size was medium (SMD = 0.46, 95% CI: 0.23–0.69). Subgroup analysis found that some groups to have better effectiveness. Specifically, the effect of STEM attitudes (SMD = 0.01) was smaller than computational thinking (SMD = 0.48). Educational robots had more significant effect on boys (MD = 0.39) than girls (MD = 0.27). The effect in primary school (SMD = 0.27) was higher than in middle school (SMD = 0.04), and the effect was great on short-term instruction with educational robots (SMD = 0.35). Based on these results, the study makes some recommendations for educators about strengthening the influence of educational robots on STEM attitudes, improving the persistence of their learning effects, and further exploring their application models.

No abstract