Page 78 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education

76
Chapter 3
3.5 Conclusion and Discussion
As to the first research question — How can the intended learning outcomes of Computational Science (modeling and simulation) be described in operational terms — we have obtained an operational description based on literature on modeling and simulation. The elements of the description turned out to be suitable to classify simulation modeling activities of the students in our study. However, some of these had to be grouped together since the separate elements could not be distinguished. The resulting operationalization contains the elements purpose, research, abstraction, formulation, requirements/specification, implementation, verification/validation, experiment, analysis, reflection. This ‘blurring’ of activities is also described by Wilensky and Rand (2015).
In answering our second research question — What data sources are suitable to monitor students’ learning outcomes when engaging in modeling activities — we found that every source enabled us to observe some aspects of the modeling process. The interviews provided the opportunity to observe all the aspect of the modeling process, closely followed by recordings of students at work. In the project documentation, the description of the model and the reflection are well represented and experimenting and analysis not so: contrary to the presentations, where exactly the opposite happens. The surveys, in their present form, did not provide much insight into the modeling stages the students engage in.
We are planning to use our results to develop an assessment instrument. In order for such an instrument to be feasible for classroom usage, a combination of project documentation and class presentation are promising data sources that enabled us to capture all aspects of students’ work. Our findings suggest that the instructions for documentation and presentation could be sharpened to improve visibility of (systematic) experimentation and data analysis within the model.
Finally, in answering our third research question — What specific challenges do the students experience when engaging in modeling activities — we identified some difficulties. Many students could not decide what to model exactly, and found it hard to decide on the level of abstraction and formulate the problem suitably for modeling through ABM. While all students managed to program something, not all of them were able to program all they wanted because either they could not decide on the relevance of a feature, or they did not know enough NetLogo to