I was just in the midst of ruing the new teacher evaluation regulations in New York, which allow state test results (or value-added scores derived from them) to account for up to 40 percent of a teacher’s evaluation.
And now there’s more to rue.
According to the New York Times, New York City will be rolling out new standardized examinations, which will be used mainly to evaluate teachers, not students. This will mean one or two additional tests per year for elementary school students, and up to eight more tests a year for high school students.
It isn’t clear that these tests will affect the students in any way (except that they will take a great deal of time and test prep). But they will affect the teachers a great deal.
This part of the article has me scratching my head:
City officials want their tests to be different from the mostly multiple choice tests the state uses. A proposal given to testing companies for bids in April asks that the exams be based around tasks, like asking students to progress through a multistep math problem, modify a science experiment to get a different result, or write a persuasive essay. They should also reflect the more rigorous Common Core academic standards that New York and other states have adopted.
It’s possible that these will be an improvement over the bland and basic multiple-choice tests (which students will still have to take). But isn’t it a bit rash to use these brand-new tests to evaluate teachers, when many schools haven’t even figured out their curricula? And while these tasks look complex on the surface, is there any guarantee that they will be so in reality? Will the scoring follow a mechanical rubric, or will the scorers read students’ writing carefully for its substance?
Even more peculiar is the DOE’s assertion that the purpose of these tests should be “transparent”:
Mr. Polakow-Suransky encouraged transparency between teachers and students when they are administered for real. “I don’t think that it should be a secret that part of how teachers are evaluated is how kids’ learning goes in their class,” he said.
Transparency is laudable in principle. But what message are you sending the students when you tell them that their test scores on a particular test will affect how the teachers are evaluated? Especially if the stakes for the students are low, students may intentionally do badly on a test when they dislike the teacher. Two or three kids intentionally flubbing a test could cause a teacher’s value-added score to plummet. It’s one thing if teachers inform students that they (the teachers) are evaluated partly on the basis of the class’s performance. But the article suggests that students will be told that these very tests will be used for teacher evaluation. If that’s the idea, it should be reconsidered.
Moreover, many kids at this point are simply unused to concentrating on any task, alone, for an extended period of time. Such concentration is very important but does not come overnight. The DOE’s emphasis on group work has not helped in this regard.
What would be wrong with (a) working out curricula (not necessarily a single curriculum); (b) developing tests that correspond with these curricula; (c) trying out these tests; (d) carefully reviewing them for problems; and then (e) including them in teacher evaluations, along with other measures, if they indeed live up to their promises?
Granted, these new tests may take some of the weight off of the state tests, which have relied heavily on multiple-choice questions and low-level “constructed response” questions. This may be a good thing. But it makes little sense to give students these new untested tests, telling them that they’re meant to evaluate the teachers, and then to take the results as a reflection of the teaching.
Yes, I hear the reply: what, do you want to have this both ways? You don’t want the state tests to make up 40 percent of the evaluations, but you don’t want these new local tests, either. What do you propose?
Certain kinds of tests fit the curriculum so well that they require little if any test prep per se. I support that kind of test. But it’s hard to believe that the new tests are of this sort, when the curricula themselves are not clear yet. Nor is it clear that these new tests will have much gradation in the results. Many students may find themselves utterly unable to complete the tasks; others may complete them with ease. Will there be a relatively even distribution between the two poles? We do not know yet. Will the students have any reason to care how they do? We do not know yet.
A combination of state and local tests is preferable to state tests alone, when they complement each other. But since the local tests are new and supposedly different from what has existed before, they should be tried out and examined rigorously before they are used to evaluate teachers. They should be examined against the curricula in formation. They should have some meaning for the students. Too many steps are being skipped in the scramble for teacher evaluations.