User:Carl Boettiger/Notebook/Teaching/2009/05/14
From OpenWetWare
Teaching  Main project page Next entry 
Teaching Teachers to TeachYesterday David Harris and I joined Julia Svoboda in running a little workshop for high school and middle school teachers on using computer simulations and models to teach about evolution. Our platform of choice was netlogo, in which we had tweaked the builtin peppered moth simulation. The session lasted two hours, with only a short introduction and mostly handson exploration using the software in small groups. While this was designed as a guided exercise with a worksheet, it took on more free form in most groups, with teachers varying significantly on how closely they attempted to follow the exercise, how much they engaged their neighbor, and how much they explored independent of either. We all spent much of the time circulating and discussing with one or two individuals, which seemed valuable and was also great fun. It requires a different skill set then teaching from the front of the classroom, and one that I no doubt have much to learn about still. Working with leading questions without falling into the “guess what I’m thinking” track, exploring different outcomes and reasoning about surprising results are all part of the challenges and the fun. One interesting idea to wrestle with is that outcomes didn’t have a simple onetoone mapping to explanations. Extinction could happen for many reasons. Increasing a parameter such as pollution could have no effect for a while, and then have a sudden effect — such as population crashes. Either Faster mutation rates or stronger selection could help a population deal with a rapidly changing environment, but making either of these too fast could kill the entire population. Listing all the patterns would take some time, but they can all be understood in terms of the model. Understanding the model well empowers one to explain each of these relationships. We found a powerful test of this understanding was the ability to manipulate the set of parameters to get the desired result, or predict the outcome of a proposed manipulation. Some manipulations have surprises, which help identify where the learner’s conceptual model has left out some element. After the session, I stayed for the reception and a chance to talk more to some of the teachers. A decent but lengthy presentation given over dinner on the topic of plugin hybrid cars sidelined most of my efforts, though I did get to speak to some of the other workshop instructors who remained afterwards, including Rich and Judy, (both of whom live in Davis and offered me a ride home).
