"When is the pressure zero inside a container? Mission impossible": 7th grade students learn science by constructing computational models using the much.matter.in.motion platform

The paper explores students' learning about gases in chemistry through constructing computational models of complex systems with the new Much.Matter.in.Motion platform (MMM; [7]). The design of MMM is based on the agent-based modelling approach to complex systems. The interface is governed by an epistemological structure into which programming blocks are integrated. They are inserted into one of three sections: properties, actions and interactions, for each population of entities. 22 Seventh-grade students' learning of science and systems concepts and modeling practices using the MMM are compared with 28 students' learning with a normative curriculum. Findings shows that conceptual learning is deeper and more integrated; the system's levels are better distinguished and related. Students' modeling expresses a gradual increase in explorative-ness and most importantly, show a gradual shift from relying on the external modeling platform to activating and developing their own internal mental models on the fly.