Page 138 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education

136
Chapter 6
6.4.6
Testing the Model
on more general and not so specific assumptions. Group G2 tried to make a model for evacuation of a burning building in general, thus with any floor plan. In their initial attempts, the agents in their model were walking through the walls on their way to the nearest emergency exit and the students did not know how to program the model to alleviate this behavior. Instead, they decided to use only a particular floor plan in combination with additional measures to prevent this problem from occurring.
2. Omission occurs when students are not aware there are aspects of the phenomenon under scrutiny that are essential in order for the model to represent that phenomenon correctly and thus make the model fit for its intended purpose. An example of such an omission is given by group G5 who modeled a cheese barn. In their model, the cheeses only get the opportunity to ripen if they do not get sold first, whereas in this case it is essential to put cheeses aside and let them wait until reaching their intended age before being sold.
3. Circular reasoning is a misguided attempt by a modeler to use the outcomes of a model as model’s assumptions. Similarly to omissions, students are not aware of this logical fallacy when they employ it as a solution to the problem for which they have no idea how to solve otherwise. Group G4 unwittingly used Ohm’s law to prove it. When the teacher pointed this out, student S7 commented, “and that was purely because at the moment we had no idea how we’d calculate the resistance and it didn’t occur to us, like, you can’t use that at all” and went on to conclude, “the result is that our model works like the Ohm’s law, but it doesn’t confirm it.”
Once the assumptions are established as a result of the research and abstraction process, the model is developed and implemented, i.e., programmed. In the process of verification, the modeler ensures that the model really does what they think it is doing (Sturrock, 2015). The simple fact that a program runs, i.e. there are no syntax errors, is no proof of a correct implementation.
In this section we report on what students saw as errors themselves and in the next section we report on what techniques they employed to diagnose those errors.