Doug Lemov's field notes

Long time friend and colleague Chi Tschang forwarded me this clip about a simulator that folks at University of Central Florida built to allow teachers to practice and learn without making mistakes with real kids.

Interesting idea.  I asked Chi to pass along some more thoughts. 

First, he observed the historical results of real flight simulators, courtesty of Daniel Coyle:

“In the 1920s, airplane fatality rates were about 25 percent. In other words, fully one in every four flights led to severe injury or death. Now, those same rates are close to zero. In the past decade, there have been – on average – only two deaths for every 100 million passengers on commercial flights. In 2010 – a particularly good year for airlines – there was not one single death due to an airplane crash. What caused this remarkable improvement in airplane safety?

Coyle argues that the difference-maker was radical improvements in how we trained our pilots. In the 1920s, a young New York inventor named Edwin Link built the first flight simulator, which recreated the sensations of being up in the air so well that his initial market was amusement parks. Now, it is a given that aspiring pilots use flight simulators in their training. While they don’t fully replace the actual experience of flying, simulators do allow novice pilots a safe and efficient opportunity to practice, fail, learn from errors and try again. Pilots get multiple at-bats; they can take off and land dozens of times in a few hours. Coaches can pause the simulation to take advantage of teachable moments or throw intense weather situations at the student with no risk to safety.”

Then he added some thoughts about applying to teaching:

For the past few years, I’ve wondered if I could invent the equivalent of a flight simulator for teachers, so new teachers could practice safely before going live in front of their own classrooms with all the hazards posed by having a few dozen adolescents in a confined space. I certainly wish I had a simulator prior to my own fairly disastrous, entropic first year of teaching.

It appears that a few professors at the University of Central Florida have beaten me to it. A team led by professor Lisa Dieker has created a TeachLivE lab with a “classroom simulator” used to train new teachers. Here’s how it works. Each teacher-in-training wears a headset and talks to a virtual classroom with five student “avatars.” Each avatar is programmed with a common, realistic student profile – for example, your typical overachiever (sharpened pencils on desks) or your chronic discipline concern (slouching, calling out). The virtual classroom is designed to be interactive: for instance, avatars roll their eyes if the teacher has poor classroom management skills.

I’ve never seen this program in person but it appears to confer all the same advantages of a flight simulator. For one, safety. Dieker said, “We didn’t put any kids at risk while our teachers learn.” For another, efficiency. Researchers estimate that spending 10 minutes in the simulator translates into one hour of live classroom experience. Another benefit is data collection. All sorts of helpful data –  teachers’ responses to misbehavior, teacher-to-student talk ratios, etc. – are collected by the program’s software.

Neat idea; neat analysis. 

Though it raises a question for me (Doug talking now, not Chi): How is the simulator more effective than simple role playing and practice with other adults?  The article notes, for example, that Dieker needs to get the lesson plan a few weeks in advance so she can program the simulator for how to behave.  Sounds like a high transaction cost and long lead time to be able to adapt the simulator.  The ironic thing is that many of us use a flight simulator now, and have been for years (Chi included!)… at high performing schools like Chi’s and our Uncommon Schools we get all of our teachers together, have them play the role of students and ask someone to “teach” in front of them.  Like with a simulator our folks improve without the risk to kids.  But our simulator is really smart… ie the “students” are role playing teachers who can give really good feedback to the teacher.  Plus as a simulator a “group of role palying teachers can be re-programmed really quickly and is “smart” in adapting to the teacher.

So: 1) I’m with Chi completely that simulation is key to increasing success by orders of magnitdue without damage.  and I think the simulator is a cool idea.  I can see ways that it could allow people who don’t have the benfit of really smart and willing colleagues to practice.  But I’m also a Luddite … I wonder if collegial role-playing as a form of practice is a stronger (and cheaper tool).

So what do you think?  Will grad schools all have simulators 10 years from now???

Flight Simulator; Chi Tschang