Sunday, January 07, 2007

Merit Pay For Teachers

Joanne (see blogroll at left) has a post about a University of Florida study on merit pay, and for the conclusions drawn I can only say "duh."

Really, what did they expect to find?

6 comments:

Anonymous said...

Well OK. Paying big bucks for higher test scores raises test scores. This makes sense. Duh indeed. What it does not address is whether the kids are learning much of anything beyond how to pass a specific standardized, high stakes test.

If my district would pay me an additional $15,000 for increased test scores, I'd have a bit of a problem. In my English department, about 93% of kids pass the TAKS test (we're in Texas). It would be extraordinarily difficult to raise that passing rate much. However, if I did little but relentlessly drill the kids on the tricks--and I do mean tricks--necessary to pass and excel on the test, I suppose we could get some very modest increases, and I would get some money.

The question remains: what, beyond learning the tricks necessary to pass the test, are the kids learning under such a merit pay system?

It's not hard to envision a situation where excellent teachers are already maxed out in terms of their ability to better student test scores but a few average teachers might be able to do better and become better teachers. They, not the people who have been for years consistently excellent, would reap the financial rewards. Is that what we want?

Darren said...

I think the idea here is to identify these exceptional teachers and use the pay incentive to get them to move to schools where they *can* improve not only test scores, but also learning.

You seem to have the belief that test scores do not indicate learning. I don't share that belief. Tests which, psychometrically speaking, are both valid and reliable--and it's not so hard to create tests that are both--can be a strong indicator of learning. I find it odd that teachers, who give tests all the time, and who probably have no training is psychometrics at all, denigrate tests. Odd indeed.

Anonymous said...

Dear Darren:

I've no doubt that test scores can be an indicator of knowledge gained, or learning, if you will, if the test is valid and reliable. However, a great many standardized tests are not. In the experience of Texas teachers, the TAKS test, particularly for English, is eloquent testimony to this limitation. About 2/3 of the test is writing which is graded by people paid to read hundreds, perhaps thousands of essays and apply a rubric. Thus, the process is completely subjective for most of the test and the most important part of the test. Fail that primary writing section and it is impossible to pass the test. In addition, if the prompt is poorly constructed or confusing, and many are, one gaurantees poor results.

Those who construct and grade the tests will of course tell you that despite being subjective they are absolutely reliable as they have worked all potential human subjectivity out through their brilliant methods and processes. Yeah. Sure.

And while I have indeed studied psychometrics, one need not be expert therein to understand that tests have, by their very nature, limited utility and can and should make up only a part of a competent's educator's understanding of any student's progress. Unless we are willing to concede that the score of a single test can tell us all we need to know about any individual and should be the sole determinant of graduation, we can see the folly in relying overmuch on test, particularly the score of any single test.

I "denigrate" tests only when they are used beyond what experience and reason tell us they should be used for. I denigrate them when they become the be all and end all in the educational process, and I denigrate them when one must spend so much class time preparing for them that they displace learning that is far more valuable, meaningful and lasting for society and for the individual student.

Tests have their place and any competent educator uses them with a clear understanding of their strengths and limitations. What is odd is that some--and I'm not accusing you of this--put all of their educational eggs in a single testing basket despite knowing that this is not valid, nor does it yield reliable results.

Darren said...

You're right that all tests aren't valid and reliable--and that's where part of the problem lies. The other part is when tests are used for things beyond what they're designed for. Those two are rather easily fixed--seriously--if there's a will to do so. Unfortunately, inject a little politics and all too often the necessary will goes away, replaced by turf battles, paybacks, etc.

The former Kimberly Swygert, and her blog Number 2 Pencil, were tremendous sources of information about the capabilities and limitations of testing. I guess it's good that she got married and gave up her blog, but I miss her knowledge on the topic.

Anonymous said...

Sure merit pay will work.

I mean, check out those P.E. teacher "test scores" we need to judge. Or special ed kids. Or Advanced Placement kids. Or ESL kids.

Using test scores is the politically correct way to judge who deserves merit pay. It shows nothing of true teaching. All a parent needs to do is sign their kids out of testing and that decides whether or not a teacher gets a pay raise? Please.

By the way, I teach Government, Economics, and International Studies to Seniors. No Seniors are tested. No pay raise for me?

Politics out of education sounds good right about now.

Anonymous said...

How about politics out of education, not to mention using fuzzy nonsense instead of tests to evaluate students only after you stop dumping students who can't add or spell in the university classroom?

Until then, raise the standards, and test ten times as much as you're testing now. And merit pay? Sure.