Enhancing computational thinking in early childhood: A cluster-randomized controlled trial of interactive STEM app intervention and the role of engagement

This cluster-randomized controlled trial investigated whether a 12-week interactive STEM app intervention enhances computational thinking skills among young children aged 4–6 years and whether app usage engagement predicts learning outcomes. Ten classrooms from five kindergartens in Northern China were randomly assigned within each kindergarten to an intervention condition (5 classrooms, 125 children) that received structured app-based activities three times weekly or a waitlist control condition (5 classrooms, 125 children) that continued with traditional educational practices. Computational thinking was assessed at pretest and posttest using the TechCheck-K, a validated assessment designed for young children that does not require prior coding knowledge. App usage engagement was quantified through a multidimensional composite index capturing task completion, time on task, persistence, help-seeking efficiency, navigation independence, self-correction, and attendance. After controlling for baseline ability and demographics, children in the intervention group demonstrated significantly greater computational thinking gains than controls, with a small-to-moderate effect size of Cohen’s d = 0.39. Within the intervention group, higher engagement significantly predicted superior posttest performance, although the unique variance explained was modest (approximately 1.5% beyond baseline ability and demographic characteristics) and the relationship is specific to the intervention context. These findings provide evidence that interactive STEM apps can enhance early computational thinking development when implemented systematically, and the main intervention effect was robust to mixed-effects sensitivity analyses accounting for clustering at both the kindergarten and classroom levels. Engagement emerged as a statistically significant, though practically modest, predictor of technology-based learning outcomes within the intervention context, underscoring the potential importance of fostering active, persistent interaction with educational technologies without overstating its standalone contribution. Implications for early childhood education practice, app design, and technology integration are discussed.