The effects of educational robotics on students’ learning outcomes: A three-level meta-analysis

Abstract

Educational robotics has been extensively employed to support students' learning. Although the effect of educational robotics on students' learning has been widely examined, extant empirical studies have reported contradictory conclusions. A comprehensive meta-analytic study was conducted to examine the effects of educational robotics on students' learning outcomes across various educational levels and school subjects and explore potential application-related, interaction-related, research-related moderators of the effects. A total of 108 effect sizes from 58 studies (N = 5806) that met the eligibility criteria were included in the meta-analysis. The results of the random-effects model indicated a moderate-to-large overall effect (g = .756) of educational robotics interventions on students’ learning outcomes. Specifically, educational robotics had a moderate-to-large impact on cognitive learning outcomes (g = .738) and a large effect on affective learning outcomes (g = .808). The moderator analysis revealed that control group condition was the only statistically significant moderator, indicating that the relative advantage of robotics is larger when compared to traditional instruction than to other technology. The non-significance of most moderators suggests the robust and versatile applicability of educational robotics across diverse educational contexts.