Mr. [Arne] Duncan joined Ms. [Michaelle] Rhee in advocating the use of student test scores as a measure of teaching ability and paying teachers for performance. Ms. [Randi ] Weingarten agreed that teacher performance should be measured, but objected to the recent publication of teacher evaluations by the Los Angeles Times, calling that particular evaluation system flawed. Mr. Duncan, in contrast, praised the publication.
Meanwhile, teacher union leaders in Los Angeles have urged a boycott of the Times and asked union members to suspend their subscriptions. So much for thoughtful discourse.
On this issue Ms. Weingarten and the teacher unions are fighting a rearguard action. A recent Gallup poll found, unsurprisingly, that 72 percent of public school parents believe teacher pay should be based on performance.
That’s a reasonable expectation, since the future of their children is at stake.
The question, of course, is how to measure performance. Raises and firings used to be based on subjective evaluations by school principals. Surprise, surprise: it turned out that teachers who acted on their own skilled notion of what was best for their students did worse than those who brown-nosed the principal. Bad schools got worse. Back to the drawing board.
Some districts with “objective” evaluations have tied bonuses and even retention to absolute pass rates on standardized tests. (Let us leave aside for the moment the question of whether these tests actually measure anything useful about student success, and the additional question of whether “teaching for the test” ignores other important educational values.) Surprise, surprise: the teachers in schools with at-risk students do worse than those in schools with more fortunate students, so any teacher who can leave gets out of those schools, making them even worse. Back to the drawing board.
The latest trend is “value added” evaluation, basically seeing whether a teacher’s students do better on the standardized tests this year than they did last year. A key problem with this is that extreme scores tend to fall back towards the middle, so a teacher who gets a class with students who did very poorly last year is more likely to show “value added” than one who gets a class with those who did very well. For example, someone teaching an Advanced Placement English class of students who all got 100% on the test last year cannot possibly show any added value, while one who gets a class of at-risk kids who all failed can hardly help but to show some improvement. So, surprise, surprise: the teachers who teach the students with the highest scores get the smallest raises and are the most likely to be laid off.
The same unintended and unearned consequences under “value added” follow for teachers who happen to get an unusual number of disruptive students who adversely affect their classmates, as opposed to teachers who get an unusual number of nurturing students who help their classmates. The [i]Los Angeles Times[/i] rankings were apparently based on the “value added” for students who were in a teacher’s home room class at the beginning of the year. There was no allowance for the fact that many students transferred to other classes during the course of the year or for the impact of all the other teachers (and other school personnel) the students saw during the course of the day. The seemingly “objective” test thus hinges on a number of random factors that have nothing to do with actual teacher performance.
The [i]Post and Courier[/i] editorial thus confuses two independent issues: “Yes, almost everyone agrees that teacher pay should be based on performance,” but “No, it is not the case that those who oppose a particular method of performance evaluation are necessarily teacher-union dupes fighting a rearguard action.” This is a complicated issue that can’t be fixed by applying some sort of mechanical formula.
The other side to this issue is the idea that parents should be privy to the performance of those who are teaching their children. As it is, a parent has precious little input in WHO is teaching her child. The LA Times has at least made it possible for some transparency in this type of situation. It seems that the Times is not the first publication in the first city to do this, just perhaps the biggest and loudest!
If we honestly believed that children’s scores on tests provided an adequate evaluation of teacher’s performance, we need to think carefully about what is happening in schools today. Teachers are already teaching children to test – that’s all they do day after day. That doesn’t make them good teachers – the kind of teachers of who have a substantial effect on the lives of their students, the teachers who teach their students to think and encourage them to exercise their creativity while at the same time insisting on intellectual discipline and appropriate social behavior. There aren’t many of those around anymore. A lot of teachers in my life taught me how to take exams – only a few taught me how to live my life.
Dale has written well here. But see my final comment on the related entry below. What parents and schools committees are looking for is an EASY way to assess performance. This is an impossible demand. Americans are impervious to this dilemma because they commonly think that ANYONE can teach if he puts his mind to it.
Incidentally, really bad teachers are easy to see. They give their incompetence away with every move and every word. They’re not the problem. Larry
The solution would be to have vouchers where parents could choose the schools that they feel best fit their child, and have the best teachers.
Oh, for the 1950s when I was in high school in Upstate New York! It was tough getting through school……and it was meant to be tough!
#1 – fair comments. There are so many variables involved it is hard to arrive at a true picture of competence. The real issues, I believe, are to do with matters schools have little or no say over: the home life and social world of children. Why, for example, are boys failing to compete in the western world? This is a no go area for political reasons.
Let me ask a simple question: Precisely what performance are you grading teachers on ? What is being evaluated?
The proximate goal is that bad teachers get fired and good teachers hired. Or do we wish to fire the mid-line teachers too? Do we wish to keep only the very best? Stop and think: Most people have no idea what “very best” means, and I am not sure I do. Until we are clear what we want from teachers (and administrators), “best” is an empty set. Are we not clear what we want? Or is it the case that we know what we want, and that is, high test scores? Is this to be the sole judge of excellence? is there anyone ready to admit this? Remember, even the “best” colleges have refused to allow the SAT scores – after all, a well designed test and of long standing – to be the sole, or even primary, determinant for acceptance. What can the reason be? Larry
Here is the major points: public schools, especially the larger metropolitan ones, stink and they are horribly inefficient.
Detroit public school education costs $11,500 per student per year. There is a 25% graduation rate and it is very questionable whether those graduates really have the basic qualifications. That cost could buy a fantastic private school education.
People object to “teaching to the tests”. If the tests test the skills that are needed: arithmetic, reading, writing, geography, history, then teaching to the tests is precisely what is needed.
There is a good article in the NYT (imagine that) about “value added testing”, [url=http://www.nytimes.com/2010/09/01/education/01teacher.html ]here[/url]. One definite objection is that teachers of gifted kids have a harder time qualifying for bonuses because the kids are already on the upper tail of the bell curve. The article cites a high school physics teacher whose kids are the brainiacs and never qualifies for the value added testing bonuses. So it is not perfect. But if one has lower performing students, who have ample room to improve, and they do worse under a teacher, that is pretty damning, because the students are in the low tail of the bell curve.
Larry writes, “Incidentally, really bad teachers are easy to see. They give their incompetence away with every move and every word. They’re not the problem.” How are they not the problem? It must be incredibly discouraging for the good teachers to get paid less than terrible teachers who happen to have been around longer because seniority is the only measure that the teacher’s unions will allow to be reflected in teachers’ pay.
Beg pardon Robroy. I meant that, being easy to spot by virtually anyone, they should be easy to get rid of. It should be easy therefore to change the rules that allow bad teachers to get canned. Now, we know it is NOT easy, but the problem is not identifying turkeys. I have tried to fire third strikes, and it can only be done over YEARS, by building up evaluation after evaluation which show that the teacher is unable to improve his skill. Some schools have tried to beat the system by firing everyone, and then hiring back piecemeal. This will work if they can get away with such a gambit and have the courage to stick to their guns.
The problem of teaching to the test is (a) the test itself and (b) the narrowness of the test which excludes all manner of important material. AS to both cases, for example, many tests simply ignore English grammar beyond the simplest material. But why? Because the English teachers who make up the tests know little about grammar themselves. And after all, when a teacher gives a test in his class, he has been “teaching to the test” all along, hasn’t he? This is fine and I believe in it.
But let’s take English composition. What test is effective? The test itself must be an essay and the effectiveness of a teacher’s work in composition cannot be demonstrated by a single essay on an assigned topic in a test room. Here one must look at essay after essay by the same student, and what system is capable of this? Larry