Attempting to teach the nuanced mechanics of dynamic systems is always a challenge. From economies to climates the number of variables that impact the observable manifestations of these systems can be overwhelming. Complicating matters is that these systems are in constant flux – they won’t stand still! Not even the most venerated climatologist or economist can say with great certainty how these systems will behave on local scales.
NetLogo is an open-source visualization tool that attempts to aid in the appreciation of such complex systems. Developed by Northwestern University, NetLogo is a freely downloadable software program that demands minimal hard drive space or processing power. NetLogo executes and animates agent-based models, that is, models in which the collective interaction of any number of agents (from two to two-thousand) generates novel outcomes through time.
Learning is an exploration. Allowing students to play with the variables that impact how systems behave can instill an appreciation for the complexity that underlies many social and physical systems. NetLogo can be used as a tool to demonstrate that much more goes into economic behavior than just supply and demand, and that climate is far more complicated than just temperature.
To walk through an example, within the Models Library we can select a model called “Wolf Sheep Predation” (under the Biology folder). This model was designed by the same Northwestern team that developed the program. All developer information, the reasoning behind the inclusion of variables, and suggestions for running the model are available under the Info tab of the user interface.
This particular model demonstrates how sheep, wolves, and grass (optionally) interact in an ecosystem under certain conditions, including how much subsistence wolves get from eating sheep, how much sheep get from eating grass, how fast grass grows back, and the reproduction rates for sheep and wolves. The effect of these conditions may be modulated by the user to see how different ecosystems that may evolve over time.
If you run the model under the default settings, you see there is a great deal of randomization built-in – running the model under the same conditions over and over produces different results. Under the default conditions there is a tendency for sheep to eventually reproduce exponentially, but this does not happen every time. Depending on how frequently the agents interact in the user interface, different ecosystems may emerge. This is a valuable lesson itself – that similar conditions do not always produce similar results. However, after altering the effect of the programmed conditions on the agents, it becomes easy to see how ecosystems are complex networks of interaction, and slight alterations of variables can have dramatic effects.
To be sure, this is a model that attempts to emulate behaviors of agents based on coded parameters. The point of using the “Wolf Sheep Predation” NetLogo model as an educational tool would not be to determinatively assert that, “when sheep and wolves interact, THIS is what happens.” Quite the opposite, the goal is to teach students how diverse and numerous are the agents that impact the phenomena we observe.
It is also worth noting that NetLogo is not a program with which many campus computers are probably pre-outfitted. For this reason, if you want students to explore the program on their own, it would require asking them to download it on a personal computer. However, if there is resistance to this, in-class lessons in which the instructor demonstrates the emergence of complex interactions can also be beneficial.
Below are just a few of the many tutorials and classroom resources for implementing NetLogo:
Included here is a Sample Assignment utilizing the above Wolf Sheep Predation model: