“Scratch”-ing computational thinking with Arduino: A meta-analysis

•Current meta-analysis provides evidence for efficacy of Arduino and Scratch.•Arduino and Scratch foster students’ computational thinking (CT) skills.•Students’ grade and duration of interventions moderate effects.•Arduino and Scratch provide low cost hands-on access to physical computing.•Hands-on Arduino- and Scratch-enabled activities have positive effect on CT skills. Empirical research into the efficacy of Arduino- and Scratch-based interventions in science, technology, engineering, and mathematics (STEM) has greatly increased in the recent years. The purpose of this meta-analysis is to compute the overall effect of these interventions. The settings for the interventions being meta-analyzed are K–12 and post-secondary classrooms. This study uses a meta-analytic methodology to aggregate the effects of 11 quantitative studies with 29 unique effect sizes to produce a cumulative effect size along with relevant visual presentations of the effectiveness of the interventions. Data for this study were collected through a detailed literature search using the PRISMA guidelines and all analyses were conducted using STATA 16 statistical software. The results indicate that Arduino- and Scratch-enabled interventions had an overall positive effect (d = 0.67 (CI: 0.40, 0.95) on students’ STEM academic achievement and their perceptions towards STEM. Hands-on engineering activities that encourage computational thinking and employ a combination of Arduino and Scratch can have profound impacts on students’ problem-solving skills, understanding of computation, and interest towards engineering professions. This research informs STEM instructional leaders and policy makers about the positive effects and future promise of open source Arduino hardware and freely available Scratch software in STEM education.