The model types that would work best with ADI activities include the constraint system (which sounds a lot like the scientific method), the trend game, in which one can analyze trends to make predictions, and aggregate behavior models which assume random motion of particles or agents. This is because these models are very useful for designing experiments, and experiment design is one of the central experiences in an ADI activity. Epistemic forms that do not lend well to ADI include lists, hierarchy models, and compare and contrast models. The latter games are better suited for open-ended,exploratory inquiry, and do not pair very well with experimentation.
One game that would work well with an ADI activity is the “trend game” – here, the classic predator-prey relationship can be explored. The questions for the ADI activity would be something like, “What is the relationship between the wolf population and the deer population? What happens to the wolf population when the deer population decreases?” Students would be able to model this relationship in starlogo nova and make predictions based on the model. They can conduct experiments with the model, They would also use the internet to research aspects of predator-prey relationships.
Another model that would work well for ADI would be the aggregate behavior model. In this model, all particles are assumed to move randomly and interact with each other in different ways under different conditions. The question for an ADI activity would be “How does osmosis work?” Students would be able to model the movement of water molecules across membranes either with starlogo nova or with a physical model like a glass of water and a semi-permeable membrane like a plastic bag.