A model for modelling

The modelling curriculum of the environmental sciences bachelor in D-USYS is currently fragmented. This results in alumni with a large variation in understanding of both practice and theory of modelling as a tools for environmental system science. This ranges from model methods and datasets to programming and workflow skills to a sense of the questions that can (not) be answered with models.

We create a template for a coherent curriculum for modelling within the Environmental System Sciences bachelor program. This starts with a review of modelling competences and learning objectives in literature and an inventory of the current course contents and learning goals. We refine these based on interviews with lecturers, other scientists, and students.

We create a special interest group «A Model for Modelling» and hold two workshops with its members to build a shared understanding among USYS lecturers of what we need to teach and how to improve our courses. These lecturers then work in peer-support groups to adjust content and upgrade didactic methods of their courses.

We also compile an inventory of common misconceptions about modelling among students from the aforementioned interviews, and use this to create a modelling concept inventory to assess students› progress on the revised learning goals.

We have an excellent chance of success because USYS lecturers cover all aspects of modelling, already know each other and know how to work in interdisciplinary settings. When the project is finished, our students will gain a better understanding and practical skills at modelling. The template and modelling concept inventory can hopefully be applied in other departments as well.

Factor 1: Even those students who never build their own models will encounter modelling results over their careers. In order to judge the value of such results, they must therefore be familiar with the basic principles of modelling. By combining different views (staff and alumni) when constructing the diagnostic tool (in the Modelling Concept Inventory), we will be able to detect whether or not those basic competences can be acquired in our curriculum.

Factor 2: The proposal is aimed at D-USYS at ETHZ, but much of the shared understanding of modelling will apply to other departments in ETHZ and beyond. If the process as such works, it can be used in other contexts, e.g. to build up a more detailed catalogue of learning objectives and competences for critical thinking. The modelling concept inventory can similarly be used elsewhere.

Factor 3: D-USYS develops and uses a broad range of models and workflows, and has staff who deal with multidisciplinary work and epistemological aspects. We can therefore do this project with people who already know each other and are used to looking beyond their own field.

We will implement a curriculum redesign that involves not only modellers themselves but also the research institutes that take on our students, and alumni who understand what the job market asks of them. We will find a way to teach a diversely used methodology (with the same core elements) across different courses. In the process, we develop a set of multidisciplinary learning objectives and a novel modelling concept inventory as a diagnostic tool.

The project has achieved the intended stronger networking between lecturers in the field of modelling and we were able to build on this in the subsequent Data Science Initiative. The MCI as a diagnostic tool still leaves room for improvement.

Students found the competency descriptions helpful in determining their level of conceptual understanding

• Make use of our table 2 of our publication to define modelling learning objectives for your course
• Make sure to coordinate the learning objectives with the lecturers who teach before or after you in the curriculum.
• Define course speficic operational tasks that correspond to the cognitive level of your learning goals, then do pre/post tests to check their validity

• Learning & Teaching Fair 2018
• Van Vliet, O.P.R, Brändle, U. „Measuring modelling “ ETH Learning and Teaching Journal, Vol 2, No 1, 2020, p.30-48