Page 143 - Getting the Picture Modeling and Simulation in Secondary Computer Science Education

Students’ Understanding and Difficulties
6.4.7 Reflection
The validity of a model is related to its intended purpose. The modelers make subjective judgements expressing how plausible, accurate and credible they perceive their models and whether they are satisfied with them.
While some students go to great lengths to ensure their models are built upon realistic data and sufficiently calibrated, checking their models’ outcomes against realistic data remains a problem.
Groups G1, G5 and G6 are aware of this issue and realize that a model can be valid without being accurate (Schmid, 2005) and that model’s plausibility is substantiated when its outcomes do not contradict previously verified theories (Bungartz et al., 2014). Student S2 said they were satisfied with their model, “when there was a visible difference between a higher density value or a higher contagiousness value. First there was no visible difference. When there was somewhat bigger visible difference, [...] we knew that it was more realistic than at first.” Group G5 was unable to validate their model against real data, but observed the outcomes of their model instead and concluded their model was valid when the trends in the output data resembled what they would expect to happen in the reality. Similarly, Student S11 said, “Because you can’t really with numbers — I haven’t checked with numbers if it worked. But rather, what do I see, does it somewhat correspond with what I’d expect in reality.”
All other groups, except G6, commented on the plausibility of their models 6 in light of possible contradiction with previously verified theories or unnatural
behavior. Student S3 said they were satisfied with their model,” when, to begin
with, as many people as possible left the building and it happened within realistic
time frame — not like initially spending ten minutes in some corner walking to and fro.” Group G3 was satisfied when their model’s outcome was that the number of people in their sustainable Mars colony turned out to be relatively stable at around 17, without extreme oscillations.
On the other hand, student S7 has doubts about their model and any other model of electrons as well, “well, what is bothering me, every model is actually wrong in my view”, but still goes on to say, “yet it was a good model, because, at least it showed, what we could see was that the values it produced corresponded rather well, so then we thought, we’ll use it.”
All the students commented on the confidence in their models — i.e., credibility, and satisfaction that the desired results were achieved sufficiently. Four groups found their models reasonably credible and were satisfied, albeit with the necessary, yet unspoken, reservations in the light of the context where their models were made
141