Teaching with Netlogo

Originally, I was not scheduled to teach in the spring, but during scheduling it became clear that there was a large unmet need and was... persuaded to open a section of the writing class. For my theme this semester, I'm asking the students to explore agent-based modelling with Netlogo. I've wanted to get back to working with Netlogo and in the fall a faculty member at WPI contacted me about working with her on Netlogo. I've reconnected a bit with the community, updated and posted my models at the Modeling Commons.

My goal for their final projects is for them to identify something in the literature for which no agent-based model exists in Netlogo and to create the model and write a report about it. Some of the students are anxious about the programming, but I think it will be very doable and have laid out a series of steps to get them up-to-speed.

For the first class meeting, I had the students observe a live ant in a petri dish to write about it and then look at the Basic Ant model I constructed. In the model, the ant moves forward until it encounters the "dish", backs up, turns 20 degrees, and moves forward again. I asked the students to suggest how actual ant behavior differed from the model. One student pointed out that the model always turned in one direction while the ant sometimes turned either way. I then pointed them at a publication that shows that ants have a left-ward turning bias and we elaborated the model with a bit more logic.

For the Methods Project, I'm asking them create a multi-panel scientific figure that shows a biological pattern generated using an existing Netlogo model coupled with a photograph that shows the same pattern in real life. We talked about some of the resources available on campus (the greenhouses and collections) and I'm looking forward to seeing what they come up with. It's always one of my favorite projects.

Last semester, I experimented with trying to work with larger student groups. I've generally avoided groups larger than three, but several faculty in the department routinely have students work in groups of 5 or more. Since I only had 5 Galileo computers, I decided to try groups of up-to 5. I would call the experiment a failure and consistent with my previous experiences. As groups grow beyond three it becomes much more likely that someone in the group can't find any meaningful way to contribute to the project.

This semester, I'm debating whether to have students work in groups at all. Or maybe work in pairs. I still have a few weeks before I have to make up my mind.