Designing MoDa: A New Tool for Science Learning Through Data-Based Computational Modeling

This paper presents the development of MoDa (Modeling with Data), a free, web-based, open-source platform that integrates agent-based computational modeling with real-world data analysis for middle school science education. MoDa was co-designed with teachers over four years, to support authentic scientific inquiry by enabling students to construct, test, and refine their own models using domain-specific, block-based programming. The design was grounded in three high-level conjectures, that science learning can be supported through: representing and distinguishing students’ existing ideas with simple, shared blocks; refining models through comparison with real-world data, and linking scientific ideas to computational practices through modeling. We employed conjecture mapping to synthesize across multiple studies to trace the relationship between these conjectures, design features, mediating processes, and learning outcomes. Implementation in middle school classrooms revealed four key mediating processes: expressing diverse ideas, exploring scientific concepts through code, validating models with data, and engaging in whole-class discourse. We observed measurable gains in students’ scientific, computational, and modeling skills. This paper contributes practical insights into how educational technologies can be intentionally designed to align with learners’ intuitive reasoning and classroom needs, offering a model for integrating data-driven modeling into science curricula.