Conversing, Crafting, Creating: How Tailored ChatGPT Chatbots and Makerspace Environments Spark Innovation in Elementary Students

Abstract

Innovation skills are increasingly prioritized in primary education, yet empirical evidence on how emerging technologies cultivate these skills remains limited, especially in non-Western contexts. This study examined the separate and combined effects of two interventions—tailored ChatGPT scaffolding and a curriculum-integrated makerspace—on elementary students’ innovation. Using a post-test-only, 2 × 2 factorial, true-experimental design, 120 sixth-grade students (30 per cell) were selected by simple random sampling from four public schools and assigned to one of four instructional conditions: traditional teacher-led classroom, ChatGPT-enhanced classroom, makerspace only, or ChatGPT + makerspace. Over 12 weeks, all groups completed 15 STEM design challenges aligned with national standards; experimental groups received their designated supports. Outcomes were measured with the Young Innovators Scale (α = .92) and an expert-rated prototype task (ICC = .88). Multivariate analysis revealed significant main effects for ChatGPT, Wilks’ Λ = .27, F(2, 115) = 155.60, p < .001, partial η² = .73, and for the makerspace, Wilks’ Λ = .17, F(2, 115) = 277.80, p < .001, partial η² = .83. Univariate ANOVAs indicated large gains for ChatGPT (η² ≈ .50) and the makerspace (η² ≈ .71) on both innovation self-efficacy and prototype quality. The ChatGPT × Makerspace interaction was non-significant, suggesting additive rather than multiplicative effects. Findings substantiate the independent value of dialogic AI scaffolding and tool-rich fabrication in fostering elementary innovation and provide a data-driven foundation for integrating these technologies within STEM curricula in Arabic-speaking settings. Future research should explore longitudinal impacts and real-time AI–fabrication integrations to test potential synergies over extended durations.