Sunday, March 27, 2016
I thought this week's paper is a good extension that summarizes the different types of modeling activity that we could do when doing ADI with students. Two of them that really applies well to ADI are aggregate-behavior and trend/cyclical analysis. Aggregate-behavior model involves analyzing emergent behavior arising from individual players that more or less act independently according to a set of internal rules. This is suitable to analyze physical phenomena such as the behavior of gases, which is affected the the actions of individual molecules. It ca also be used for biology/ecology, such as predator/prey and disease spread models that we have seen. For a class using ADI, I think using agent-based modeling using computer programming is very suitable, since students could focus at first on coming up with rules and constraints that govern the behavior of individual elements, and then see if a desired behavior emerges from those factors. A trend/cyclical analysis is more often seen in history and economics, where a set of variables is analyzed over time. I think it's useful in ADI because of its focus on prediction and extrapolation of data. I think student would benefit more in the math department if we use trend/cyclical analysis. I don't think any particular epistemic game is unsuitable for ADI, because all of them require logical thinking, data analysis, prediction, and modeling. Even the two I liked the most can be seen as interconnected, because trend analysis certainly can be derived from aggregate-behavior analysis.