Modeling Goals & Purposes

Van der Valk, T., Van Driel, J.H., and De Vos, W. (2007). Common characteristics of models in present-day scientific practice. Research in Science Education 37, 469-488. Classroom modeling practices should promote understanding of how scientific knowledge is constructed and used, thus refocusing school science away from its traditional sole emphasis on acquisition of content knowledge. This paper aims to characterize the models that scientists use in their practice, to provide a starting point for the design of teaching and learning activities aligned with the current views of science education. The authors of this study developed and deployed a questionnaire asking scientists across multiple disciplines to evaluate statements about 8 common features of scientific models, and to provide additional comments. Quantitative and qualitative analysis of responses given by 24 individual scientists provided insights into how practitioners view models and modeling. A model in science is always related to a research target (which can be an object, a phenomenon, an event, a process, a system, or an idea) and is designed for a specific purpose. While a model is not an exact copy, or replica, of the target, it is usually a simplification that bears some analogies to the target and, depending on its purpose, may include some but not all of the target’s elements, thus making the target accessible for study. To optimally address their research questions, scientists must choose carefully and appropriately the type of models they use, and will often need some creativity to adapt models to their purposes. Models are used as research tools to obtain information about targets which cannot be easily observed or measured; typically, a model is used to make predictions or to explain, but also to simulate, visualize, or experiment. A model that represents current scientific knowledge about a target can also be used to facilitate making decisions about an issue (social, environmental, etc.). A model can be optimal for its target, but is usually only one of multiple possible consensus models for that target. Finally, just as scientific knowledge is constantly growing and evolving, models can grow and change through an iterative process of refinement that results from gaining new knowledge about the target; in turn, a revised or refined model may serve to generate new hypotheses and questions.

Gouvea, J., and Passmore, C. (2017). “Models of” versus “models for”. Science & Education 26, 49-63. This article presents an “agent-based account” of modeling in the science classroom, which posits that models cannot be only interpreted and understood based on what they represent (“models of”), but it is critical to consider their purpose and intended audience (“models for”). This distinction is very relevant in the context of science education reform, mapping directly to the recommendations in the Framework for K-12 Science Education and by Vision and Change. The traditional educational approach that asked students to “reproduce from authority” is being replaced by a social constructivist approach that encourages learners to have agency and to develop and use models as a way of generating their knowledge. While all models, technically, are models “of” something, the authors emphasize that the richer potential for using models in the classroom, beyond their basic representational nature, rests within their purpose (models as tools that support inquiry, exploration and knowledge construction). Model-based learning at its best should reflect authentic scientific modeling: a generative, rather than merely descriptive practice. As learners use models to explain and predict, they generate new understandings.

Schwarz, C.V., Reiser, B.J., Davis, E.A., Kenyon, L., Achér, A., Fortus, D., Shwartz, Y., Hug, B., and Krajcik, J. (2009). Developing a learning progression for scientific modeling: Making scientific modeling accessible and meaningful for learners. Journal of Research in Science Teaching 46, 632-654. This seminal study in the field of model-based science teaching and learning describes the development of a learning progression for modeling in the 5th and 6th grade science classrooms. The authors make a compelling argument, grounded in theory and evidence, that to make the practice of scientific modeling comprehensible and meaningful for science learners, it is important to clearly articulate the goals of modeling and the optimal sequence of learning experiences that students should engage with. The study was grounded within the principles that (a) the goals of modeling in educational settings should be aligned with the goals of modeling in authentic scientific practice), and (b) students should not merely interact with provided models, but generate, evaluate, revise, and use their own models. The elementary and middle-school students in the study advanced along the levels of the learning progression, by successively demonstrating the ability to (a) construct explanatory models consistent with evidence and science theory, (b) use their models to generate predictions for new, closely related situations, (c) compare and evaluate different models of the same phenomena, and generate consensus models, and (d) revise their models to incorporate new knowledge and understanding, or to clarify and sharpen the model purpose.

Svoboda J and Passmore C (2013). The strategies of modeling in biology education. Science & Education 22: 119-142. In this article, the authors elaborate on a framework developed by philosopher of biology Jay Odenbaugh to describe scientific modeling in a way that can support instructors in making informed choices about how and why to incorporate modeling in their classrooms. The authors draw particular attention to the context of modeling. In scientific practice, there is a myriad of model types, each optimal for a given purpose and context. For example, the models created and used by molecular biologists and ecologists are completely different – they have different purposes and very different levels of realism, precision, and generality. It is important, therefore, to be aware of such diversity and to identify the specific purpose and context of modeling activities. Furthermore, this article highlights some important goals of models, which are often neglected compared to those of making predictions and explanations. These less cited but nevertheless important roles are that: (a) models are simplifications of complex systems which, by virtue of their simplicity, allow scientists (and learners) to reason about very complex phenomena (providing a valuable “cognitive payoff” while learning); (b) in science, model building allows exploring alternative conceptions or hypotheses about how something may work; this exploratory power of models may be fruitfully used in the classroom, especially if learners have the opportunity to test alternate models they develop; (c) model construction or analysis often allow initially “fuzzy” concepts to crystallize or to be clarified.

Gobert, J., and Pallant, A. (2004). Fostering students’ epistemologies of models via authentic model-based tasks. Journal of Science Education and Technology 13, 7-22. Although research is still ongoing, there seems to be consensus that students’ epistemologies (their beliefs about how scientific knowledge is constructed) and their understanding of scientific content are deeply intertwined. In this article, the authors argue that appropriately scaffolded model-based learning has a dual purpose: while promoting students’ understanding of a specific subject matter, it also influences their understanding of how scientific knowledge is constructed. The paper describes implementation of a model-based teaching and learning unit on plate tectonics in multiple US middle and high school science classrooms. Data from over 1000 students were collected in the context of these classrooms, as students engaged in rich model building, explanation, evaluation and revision tasks. The results of this study indicate that students who had a richer, more sophisticated, concept of the epistemic role of models seemed to also develop a deeper understanding of the content. The authors’ concluding argument is that authentic model-based learning tasks foster in students a deep understanding of both scientific knowledge and of science as a way of knowing.