Page 108 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education
P. 108

Chapter 5
to minimize the waiting time of the customers with various needs. In line with our dedication to stimulate student engagement, the students are allowed to come up with their own research questions instead.
While these assignments allow students to make their own choices and decisions when designing their models, we needed a standard that allows educators using our assessment instrument to easily assess the quality of their students’ models. To set such a standard, for each case we constructed a minimal expert model — a description of a minimal model that fulfills the stated purpose and contains all the necessary agents with their correct behavior and interactions. Since we wanted these minimal expert models to be described on a conceptual level only, we refrained from implementing them in NetLogo because we believed that that would hinder the assessment process rather than contribute to it. Instead, for the models that our students are expected to make — two-dimensional, containing only a few types of agents, no links and no advanced behavior such as learning or sensing — we devised our own format to describe them. This format is partly narrative, borrows aspects of class diagrams from UML and exploits the idea of graphical representation of timed automata where it is possible to require that particular state transitions are allowed only under certain conditions, or only synchronously with other state transitions (Vaandrager, 2011). Here we illustrate this approach with the description of our minimal expert model of the roundabout. First of all, there is the agent type21 vehicle with its representation — inspired by UML class diagram — stating that an agent of this type has properties22 current position and target position, and behavior consisting of actions show up, wait, move and leave.
Vehicle
current position target position
show up wait move leave
Figure 3: UML class diagram for vehicle
Then there is a graphical representation of a state diagram of a vehicle (figure
4) which is interpreted as follows: during the NetLogo setup procedure, the first
21 NetLogo speaks of breeds of agents, but we use the term type to cater for those not familiar with NetLogo.
22 NetLogo speaks of agent’s own variables
    106























































































   106   107   108   109   110