Much.Matter.in.Motion: Learning Science Through Constructing Computational Models of Complex Systems

• In Event: Restructuring Concepts and Tools Through a Complexity Perspective

Abstract

Objective
The paper introduces the design of and learning research with the new Much.Matter.in.Motion platform (MMM; Levy,Saba,et.al.,2018), which enables modeling of systems in chemistry and physics. MMM makes apparent the complexity restructuration described in Wilensky (here)(Figure.2, right). The study explores how modeling with MMM supports learning science and modeling.

Perspectives
Theoretical perspectives combine constructionist views for learning, and complexity for framing the science concepts.
Constructionism proponents learning of “big ideas” through building and sharing artifacts as a central form of activity and learning (Papert,1988/93). MMM enables learning through constructing computational models using ubiquitous block-based programming (Scratch;Resnick,2003) to overcome the need to teach text-based programming.
Several innovative learning environments have been designed to help people understand complex systems, by constructing richly expressive computational models (Sherin.et.al,1993;Wilensky,1999). MMM supports students in modelling diverse phenomena using a properties-actions-interactions structure for modeling systems, highlighting commonalities in chemistry and physics by using a small set of blocks to program a wide array of phenomena. The topic of learning, gases, is restructured in the following way: both normative and restructured learning begin with particles and their behaviors; in the restructured form, these behaviors are more highly specified through their programming and emergent and stochastic processes are highlighted.
The MMM platform combines a rich NetLogo (Wilensky,1999) super-model (Saba,Hel-Or,&Levy,2019) and Blockly (Google,2012), an open-source library for creating block-based interfaces. Modeling is at the micro-level, while macro-level objects and fields are drawn in by hand. The first study used an initial version of MMM (Figure 1, left), using widgets instead of blocks, arranged and signifying the same structures.

Methods and Data Sources
22 seventh-grade students’ learning by constructing models is compared with 28 students’ learning with a normative curriculum, during six 1.5-hour sessions. They completed pre-and-post-test questionnaires, four students were interviewed. Activities were logged and screen-captured.

Results
An interaction between time and group (F(1,48)=5.09,p<0.05) shows the superior learning of the experimental group: overall, Kinetic Molecular Theory, micro-level (Table.1).
Analysis of screen-captures examined how design components, e.g. drawing the macro-level, impacted modeling. Another analysis (Figure.3) shows an earlier episode focused on micro-level behaviors and interactions and science concepts related to particles’ behavior. The later episode shows more use of science and systems concepts, including macro-level and emergent components. A greater number of models was constructed in the later episode, with more variations in the code, and thought experiments.

Scholarly Significance
Results suggest that modeling with the MMM version of the complexity restructuration, which highlights unity among chemistry and physics phenomena, increases learnability of science, systems and modeling. Conceptual learning is deeper and more integrated. Systems’ levels are better distinguished and related. Students modeling expresses much explorative-ness and the most importantly, the gradual shift from relying on the modeling platform to their own mental models.

Authors

• Janan Saba, University of Haifa

• Hagit Hel-Or, University of Haifa

• Sharona T. Levy, University of Haifa