Complex systems for middle school science

Vanessa Svihla, Learning Scientist, UC Berkeley

My interest in systems is two-fold: I would like to better understand student learning of complex systems; and as a learning scientist, I strive to understand, explore, and represent learning as a complex system. My experience learning and teaching with complex systems therefore spans disciplines and goals. While working towards my masters degree in structural geology, I taught geology courses to non-science majors, including topics on earth systems science, such as climate change and plate tectonics. I was frustrated by the inadequacy of the materials available to me at the time, in terms of conveying a systems perspective of these topics.

While completing coursework towards my PhD in science education, I learned about various system modeling tools, including STELLA and netlogo, and used these tools to model the course of innovations in social systems, including unintended consequences. I also participated in a graduate seminar exploring the teaching of evolution, focusing on the component understandings (e.g., deep time, variance, complexity) necessary to form an understanding of evolution.

In my research as a Learning Scientist, I apply integrated methods to understand learning as a fundamentally social and contextual process. As such, I apply network analysis and multi-level regression, and integrate these with qualitative data. I constantly seek – across disciplinary boundaries-ways of representing and analyzing data with a goal of moving away from fragmented reductionism and towards systems understandings of learning processes. A current limitation – and one that is well positioned to see rapid change in the near future- is that models that move (e.g., STELLA) are poorly understood by the general educational research community and when presented in static format, do not facilitate- and may even obscure- explanatory power. Increased use of hypermedia journals will chauffeur their use.

My approach to understanding the teaching and learning of complex systems is informed by neurological explanations for different mechanisms for human category learning that predict different pathways for learning explicit, rule based phenomena and for learning complex phenomena (Ashby & Maddox, 2005). This same line of reasoning can be seen in research exploring ontological differences across direct and emergent processes (Chi, 2005).

In my current position as a post doctoral scholar, I am designing and researching the impact of curricula related to global climate change and consulting on a project on plate tectonics. These curricula are intended for a middle school (~12 years old) audience, and generally taught by generalist teacher, not by a science specialist teacher. Additionally, these curricula are a part of an NSF-funded investigation on cumulative learning, and therefore are organized by a core idea of energy. The plate tectonics project teaches students about mantle convection as a driving force behind plate tectonics- with no mention of slab pull- and related convection of dye in water and density differences to the mantle. The global climate change project focuses on radiation and energy transformations, and uses netlogo models to allow students to observe energy transfer and transformations in a simple climate model. Model output is global temperature, and students interact with a series of models that include only one or two variables to understand how these variables relate to global temperature.

Two tools – Energy Stories and MySystem- serve as both assessments and as learning events. These embedded assessments focus on the core ideas of energy and are used across several projects that are part of the study of cumulative learning. Energy Stories ask students to synthesize their ideas about how energy is transferred and transformed. Students are asked to write about both everyday and scientific contexts. These afford students an opportunity to integrate understanding from a series of activities, and to apply this understanding to familiar experiences. MySystem is a tool that reflects Stella models, in that it asks students to show how energy flows from one object to another, but it lacks the mathematical engine that underlies STELLA, and makes STELLA such a powerful modeling tools. In this particular context, MySystem serves as a way for students to represent their understanding of a non-linear process, provided it occurs at a singular level. Current limitations are that MySystem cannot handle subsystems, making it challenging to represent many of the processes we are hoping to teach. Additionally, lacking a mathematical engine, there is no feedback to the student in terms of accounting for all flow in the system. In most cases, students create a MySystem in tandem with writing an Energy Story as a way to organize their thinking.

The current redesign of the global climate change project has focused on two aspects: explicating the component understandings and identifying the nonsalient aspects (Liu & Hmelo-Silver, 2009) that are mechanistically consequential for understanding the climate systems.

Challenges in my particular context relate to determining what constitutes an appropriate level of systems science given:

• The age and experiences of 6^th grade students
• Pressures to cover the scope/breadth of the curriculum
• Pressures to teach to California State Science Standards
• Desire to represent topics in ways that reflect disciplinary perspectives
• Integrating energy as a core idea

There may be tensions and therefore a need to consider optimization across these needs and goals.

Future goals for redesigning curriculum include integrating across curricular projects, such that more standards can be revisited/foreshadowed. This might include, for instance, integrated models of earth systems, from plate tectonics to climate change, or from photosysnthesis to climate change.

References

Ashby, F. G., & Maddox, W. T. (2005). Human Category Learning. Annual Review of Psychology, 56(1), 149-178.

Chi, M. T. H. (2005). Commonsense Conceptions of Emergent Processes: Why Some Misconceptions Are Robust. Journal of the Learning Sciences, 14(2), 161-199.

Liu, L., & Hmelo-Silver, C. (2009). Promoting Complex Systems Learning through the Use of Conceptual Representations in Hypermedia. Journal of Research in Science Teaching, 46(9), 1023-1040.