Merit Pay, Student Evaluations

In response to this post, high school teacher Craig Barker e-mails the following:

"I want all schools to be rid of the dead wood, I also want risk takers to be protected from vindictive administrators or personal vendettas...Here's the problem I have with merit pay that has never been fully explained to me. I teach five classes a day, in the fall for the last three years, it's been three sections of freshman world history and two sections of AP U.S. History. I have a measurable test with the AP kids to know if they achieved as I had hoped, I have the AP test. But what about those world history kids? At the semester break, they will be shuffled to other world history teachers, and I will get two new classes worth as I also pick up my International Relations class. How exactly do we measure my achievement as a teacher. Two of my three subjects are in areas that draw the brightest and best of the talent pool. Do I get a monetary reward for teaching kids who likely didn't really need my help as much as a kid achieving at the middle of the percentile scale? And when the kids do take the Michigan Education Assessment Program (MEAP) test in their junior year, how do we know that it was my effort that put them over the top? How do we know I didn't damage them and another teacher is the one who helped them? There's no way of knowing and that is why I am distrustful of merit pay, no one has ever been able to explain to me how it would work in such a way that I would know it was my efforts."

On a somewhat related topic, Jonathan Dresner and Sharon Howard are discussing grade inflation and student evaluations. This quote from Dresner seemed especially insightful:

"One thought which didn't make it into the article is that student evaluations and student learning assessments both assume that learning is a short-term process, that a student can judge at the end of a semester what impact a teacher has had, and that what a student learns over a semester is best evaluated at the end of the semester. They also assume a sort of separability which is not entirely justified, either: students evaluate teachers in comparison, not in isolation, and students do not take one course at a time (and, by the way, there's no control group, and no attempt to openly discuss evaluation metrics, just a self-referential population making up their own scales). The most effective and realistic forms of assessment are going to be post-graduation tracking, long-term studies, carefully selected and analyzed qualitative and quantitative measures."

Interesting perspectives all around.