Page 118 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education

116
Chapter 5
Case and research question. While many groups — in HAVO as well as in VWO — did not engage in any form of research and relied on their existing knowledge of the phenomenon under scrutiny, we see significant differences between HAVO groups on one hand and VWO groups on the other. Among the HAVO groups, none rose above the quantitative levels of SOLO. For example, the group exploring life on Mars (case Mars 2), when answering what do they know about it, simply wrote, “that life on Mars will not happen for a long time”. However, this group made a good model and showed reasonable performance in the rest of the assignment. In the VWO groups, we saw a great variation with some groups reaching the relational level of SOLO by performing extensive research (e.g, looking up how much carbon dioxide does a person produce) and hypothesizing about possible outcomes of the model. For example, when describing what do they hope to observe form their model, the VWO group exploring the optimal strategy for potato farming (case Potato farm 2) wrote, “we hope to find out what factors help increase the profit and what factors do not influence it that much. [...] Out hypothesis is that fertilizer will really increase the profit but that pesticides have little influence on it. To prevent diseases and pests, we believe it is important to have large distance between the plants and to remove affected plants immediately.” They continue in their logbook, “We discussed what aspects of the phenomenon influence the profit. We decided not to take into account the seasons, weather and water. We did this because we look at the ideal conditions and other factors will not have a big influence.”
Designing and implementing the model. For this part of the assignment, we see the highest scores achieved and small variation among the classes. Every group managed to design a model and more than half of them arrived at a model matching or exceeding our minimal expert model. A small number of groups, while succeeding in writing some code, did not manage to write a meaningful program and subsequently failed to use it to perform an experiment. An example is the VWO group working on case Flood who stated their case and research question but failed to implement their model properly.
Validate the model. Here we see a great variation in scores and the VWO groups outperforming the HAVO groups significantly. From the HAVO groups only one reached relational level and one group did nothing meaningful to validate their model. The response from the Cheese barn 3 group is exemplary: “The agents simulate the production and sales of a cheese barn, thus the cheese production, the sale of cheese, the agents simulated the production and sale of cheese as it happens in the real world, so this corresponds well with each other.” All the VWO groups