Page 155 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education

computing in other disciplines with the goal of helping to solve problems in these other disciplines. After we have finished this work, Sengupta et al. (2018) confirmed our ideas in their recommendations where they warned against a technocentric focus on production and use of computational artifacts and argued “that computational thinking must be reconceptualized more appropriately as an intersubjective experience” — exactly as we did.
7.2.2 Students’ Understanding and Difficulties (RQ2)
In our first and fourth studies, we focused on students’ understanding and difficulties while engaging in modeling activities — Magnusson’s component M2 — and sought to answer our second research question: How can the students’ understanding of modeling activities be portrayed in terms of their requirements for learning and difficulties they encounter? In the first study (chapter 3), we looked at the specific challenges the students experience when engaging in modeling activities in all of the modeling process. In the fourth study (chapter 6) we focused on students’ challenges related to the verification and validation aspect of the modeling process only.
Our first study resulted in the qualifications of the challenges the students
faced when developing and using computational models. Our in-depth analyses
revealed that students face two types of challenges related to the two cyclic
processes contained in our framework: those related to the entire modeling cycle
and those related to formalization — i.e., the development of a computer program
— which is an element of the modeling cycle. The challenges which we found that
were related to the entire modeling cycle are those involving the context of the
discipline where the problem at hand originates. They are related to expressing the 7 problem at hand in computational terms, interpreting the computational solution
in terms of the original subject matter, and reflecting upon the whole process. The
difficulties our students faced while constructing their models were also reported
in case of students constructing mathematical models: wrong level of abstraction
and erroneous assumptions (Maaß, 2006). Not being familiar with the affordances
of the tool used to implement the model — in our case, insufficient command
of the programming language involved — is found to have a detrimental effect
on the quality of the model being produced (Bielik et al., 2021; Sins et al., 2005).
Other behaviors we identified are characteristic for the development and use of computational models: not knowing whether unexpected behavior of a model is
caused by an error or emergent behavior is typical for the development of agent-
General Conclusions and Discussion
153