Abstract
There have been a number of hypotheses that educational robots positively influence students' learning and motivation to learn. However, such claims must be based on empirical evidence, which is why the main goal of this meta-analytic study was to answer the following research question: what is the combined effect of educational robotics interventions on the post-test cognitive outcomes of elementary students compared to nonrobotics conditions? The authors searched and selected relevant research articles, which resulted in a final meta-analysis of 8 papers published from 2018 to 2022. A standardized mean difference as Hedge's g for each study was calculated, taking into account group sizes, mean post-test rates, and standard deviations. This allowed comparison of effect sizes for standard learning and educational robotics. Publication bias was also analyzed. The funnel plot, Egger's test (p = 0.104), and Begg and Mazumdar's test (p = 0.805) indicate that the publications included in the meta-analysis are not biased. According to the results of the meta-analysis, the weighted mean effect of robotics versus traditional training can be interpreted as moderate (Hedge's g = 0.64, 95% CI [0.44; 0.84]). Our findings have practical implications, providing stakeholders, such as teachers and school administrators, with validated evidence of the effectiveness of robot-based activities, which can then help them make informed decisions about whether or not to invest resources in this teaching method.