A lot of "teacher policy debate" focuses on the bottom 20% and the top 20%. A big unsolved puzzle is how the middle 60% can improve.
The data is not promising. Teachers get a little better their first 3 years, then don't seem to improve. One suspects they are improving themselves during this time, through trial and error.
PD, when measured carefully, doesn't usually pay off. Incentives, when measured carefully, don't usually pay off. That's not to say neither can work, just that typically they have not.
* * *
I got 2 basketball related emails yesterday.
From Pete (sent to an email list of guys I play pick up with on Sundays):
I have a team in the men's A league at the Oak Sq. YMCA in Brighton. The league is very competitive with a lot of ex-college players actively playing in it. A number of guys who've played with me for the past two to three years are leaving. Some going to grad school, others to revamp their old squad with other friends, some for work and to play abroad professionally. I'm actively looking to fill some spots.
From Ryan (sent to men and women I've played with on Tuesdays)
As we have spoken about, I just signed us up for a team that is going to play Tuesday nights in Cambridge. As it stands right now, I think we got a solid group of players and are going to be very competitive in this league.
I fit in fine with the second group, the solid players. Not with the first one, the ex-college and semi-pro guys.
How do I know this? It's easy. If you play basketball, it's very easy to compare yourself to others on the court, because you see them in action.
Joe covers me. I don't score. In the next game, Joe covers Fred. Fred scores a lot. Multiply by many games, I conclude that Fred is better than me at scoring.
Teachers rarely see one another in action. The typical American teacher simply does not see his colleagues teaching very often. That's because he's busy doing it himself. Or getting ready to do it. This makes comparisons hard.
* * *
This has some policy implications, I think.
Evaluation systems are comparisons. They sort. Good, medium, bad.
And it's harder to trust evaluation systems when you don't have a good "internal sense" of where how stack up compared to others, because -- while you lunch and drinks with other teachers, and sit in meetings with them, and hear scuttlebutt from kids and parents about who they like -- you rarely watch the other teachers do the act of teaching.
Joe works hard. He gets his evaluation. It says: "Joe, you are precisely average." Given normal human psychology, Joe might think his values are under-appreciated.
But Joe doesn't see the "good" teachers teach. Imagine if he watches Fred in action. He thinks "Wow, Kid A, B, and C are kids I struggled with all of last year. But in Fred's class, those 3 are really engaged." Multiply that observation by many other teachers. Joe now has a sense of how good he is. Not only would that affect his trust in the evaluation, but it also might bolster his willingness to improve, and his ability to improve (because he knows what it looks like when done well).