Peer To Peer Education (and AI)

I learned a lot about how to play trombone from John Stuck.

John was not my band director (that was Ed Frye, from whom I learned a ton, but that's a discussion for another day). John was my section leader, the guy one year ahead of me through high school. We were different people (as witnessed by the fact that I grew up to be an English teacher and he grew up to be an army tank commander), but he was a heck of a player, and I learned a lot by following his lead. One of my proud moments as a junior was screwing around in band and imitating John's style well enough that he got in trouble for my creative addition to the piece we were playing. 

Every teacher knows about the chemistry of certain individuals in a classroom. Sometimes it's negative--that class that always goes so much better if just that one student is absent. But mostly it's part of what drives the class forward. Sometimes it's as simple as plain old competitiveness; more often it's just the tone that is set. 

As much as you work to set a tone and culture for your classroom, your students are always part of the equation. They inspire each other and support each other. 

My classes depended on peer effects. Full class and small group discussion. Peer review and editing of writing. These help magnify the effect of hearing from people who are very much at the same point in their journey as you. Fellow students reveal the many possibilities for the work, and those possibilities are more powerful for coming from someone much like you. 

The effects are obvious in places like music ensembles, sports teams, school theater productions. Students challenge each other to up their game, to bring their work up to another level, to provide support on the highest tier. You strive to play well enough to keep up with your section leader. You step up to pull your weight on the team. You perform well enough to match your scene partner. 

In class discussion, I've watched students try far harder to explain themselves to each other than to me. I've watched students help each other solve writing problems. I've seen plain old rivalry push students to avoid being outdone by classmates, and I've watched students rise into leadership roles and help create a classroom that values inquiry and exploration.

I had a class years ago that was the most wide-ranging, most willing to discuss anything and everything, bouncing and building off classmates with a magical combination of push and support. My final assessments always included a giant year-encompassing essay prompt; that year it was "What is the meaning of life?" The exceptional success of that class wasn't about me-- I was doing more or less what I did every year-- but the extraordinary combination of students.

It's a throw of the dice. Most teachers have wondered what might have happened if Student X had been born a year sooner or later and ended up in Class Y. Sometimes you can help a bit, if you've got a good eye for combining students in work groups. 

But you know what won't improve the odds of this magical alchemy happening for students? Parking them in front of a screen and having them interact with an AI coach or guide or algorithm wrangler or poor excuse for "personalized" learning. 

There are many hard conversations that go with this issue. When a strong student leaves one school for another, the school they've left has lost more than just one seat occupant. A talent drain can hurt a school as well as an industry or country. But we can't possibly ask parents to pass up an opportunity for their child so that other students will benefit. Pat might have been a leader, a positive education influence on students at East Egg High, but Pat also deserves to move to an environment where the other students will push Pat. 

We know that peer effects in education are a thing; there is enough research out there to build a small house. I find it a little remarkable that while purveyors of screen-based tutoring are willing to drag out the highly dismissible two-sigma tutoring paper, nobody seems to be asking about the consequences of virtually erasing peer effects with screen-based education

Education is about relationships, and not just relationships between students and teachers or students and the material, but between students and other students. It is a human activity, best pursued by humans with humans. MacKenzie Price and her 2 Hour Learning are so proud of making learning "efficient" by reducing education to a child spending a couple of hours in front of a screen. Sure. It would be efficient to skip all the talking and kissing and other courtship stuff and just skip straight to getting married (I bet an AI could pick out a suitable mate). It would also be efficient to skip all the messy human interaction and pregnancy and stuff and just have some fertilized eggs brought to maturity in a lab. It would be supremely efficient to take all the humanity out of being human. What would be the point? 

Exceptional, awesome, inefficiently human things happen when students do their learning side by side. We'd be silly to sacrifice that for machine-directed learning.

Here is my favorite John Stuck story. John once made a wrong entrance during a halftime show with such authority that half the band followed him. I'm not sure what lesson we all learned that day, but I feel certain it was a fully human one.