Felix, you made an analogy with quantitative risk measures, and how (mis-)usage of such has over-simplified financial transactions risk. In education, it is even more egregious to rely on metrics, computer aggregated data, and a system of student critiques of teaching staff. My point is this: At least the individuals who were in charge of risk management in banking were adults. (They might not have behaved maturely, but they had the cognitive maturity to know how to). High school, or middle or elementary school students are children. As others have said, one needs to spend time as a teacher or educator or hall warden in a junior high school to realize how minors can be. They aren’t ill-willed or malicious, but they ARE still learning, and still in a nascent stage of cognitive and social development.

Let’s put that aside though. Without the passage of time, it is non-obvious whether a teacher is good or not. Very obvious measures such as teacher absentee rate exist, of course. Yet that doesn’t require sophisticated new metrics or methods to gauge accurately.

I read today a lengthy article about long-time systemic cheating in a well-known, well-respected NYC High School. That seems to contradict the basis of most of your argument.

If you see public education as a “profitable feather bed”, then you’ve clearly never tried it yourself. Give it a few years and then tell me what you think. It is the toughest job I’ve ever tried.

Besides, you CAN evaluate teacher performance, quite effectively. But your best bet is to actually look at what the teacher is doing. Too many districts don’t bother to invest the effort in that endeavor.

We have had the same school system for over eighty years but still have not concrete way to reliably judge performance. Many foreign schools are rated superior to ours and their student scores are higher in most or all categories. Why this matters when supposedly you can’t test student performance leaves me to believe American educators have found a feather bed and have snuggled down for a profitable rest. In spite of all this many competent teachers find their way in to classrooms and at least a few students actually learn something. With automation and massive job shortages we don’t need many educated individuals anyway.

Follow the money.

But I would apply exactly the same cautions to quantitative systems. A standardized test that is appropriate to judge an inner city school is the wrong target for a wealthy suburban district. A test that is challenging for College-track students is insultingly easy for those in AP-track courses. And while there are many elements of teaching that cross all subjects, I don’t see how you can possibly create truly parallel assessment standards.

School districts are hard at work gaming the mandated state testing, “teaching to the test” instead of teaching the full curriculum. (The math test, for example, ends with basic Algebra and Geometry.) One local district even eliminated History from the (middle-school?) curriculum, so they could instead spend the extra period on the subjects that are tested.

I do wish we could compare notes in more detail so that I might better understand why the system you reference was an abject failure while the system I experienced worked effectively?

Of course my pay never depended on these evaluations, and I wasn’t at all worried about my performance being graded “unsatisfactory”. I welcomed the classroom observations and learned from the feedback. Is that how your system worked?

I firmly believe that strong administration (with the authority to hire and fire) is the key to strong schools. Trying to manage a school from the outside is nearly impossible. You might have a pretty good idea whether it is successful or not, but from the outside it is very hard to know how to fix what is broken.

This is what we had in the military. It turns out that everyone is so far above average it isn’t even funny. Simply doesn’t work.

I haven’t seen a detailed proposal to evaluate teachers through student test scores (I’m no longer in the public schools), but I don’t think the concern I raise is unreasonable. Nothing I have seen described in the press suggests how to resolve it. In MA we don’t test students in every subject and we don’t test them every year. How can individual teachers possibly be evaluated on that basis?

This is why the Chicago teachers look so bad, they are arguing against positions their opponents don’t actually hold.

(1) Evaluations must be evidence-based in best practices that have been demonstrated to lead to student learning.

(2) The evidence should be collected through a series of classroom and professional observations, both announced and unannounced. During my first three years, I was observed roughly a dozen times a year by three different administrators. A few of those were announced, but the rest were either unannounced (supervisor enters along with the students at the start of the period) or 5-minute “walk throughs”.

(3) The administrators need to be given time and resources to do the job properly. They need to be trained in both classroom observation techniques and in “best practices” to emphasize. They need to give teachers regular and detailed feedback from these observations.

Standardized testing measures the students, not the teachers. If every class entered with similar skills, then standardized testing would tell you much about teacher performance. But in the real world, we INTENTIONALLY segregate strong students from the weak students. If you believe they should be expected to hit the same endpoint by the end of the year, you are kidding yourself.