Earlier this year, the Bill & Melinda Gates Foundation — one of the nation's biggest funders of school-reform projects — announced it would use the Memphis and Pittsburgh school districts, among others, as laboratories for developing “teacher effectiveness” programs using data on student achievement and teachers' classroom behaviors.
The idea is to figure out the connection between student achievement and actions of individual teachers and use the linkage to make “high-stakes” educational decisions — decisions, for example, on which teachers to fire, which to reward with merit pay or other recognition and which teaching practices to replicate.
Microsoft cofounder Bill Gates told The Wall Street Journal that he will deem the project a success if “10 years from now … we have a very different personnel system that's encouraging effectiveness [in teaching] and our spending has contributed to that.”
He went on to say that education-improvement efforts have suffered because data on teacher and school performance haven't been available. Contrast that situation, he said, to “professions like long-jump or tackling people on a football field or hitting a baseball,” where “the average ability is so much higher today because there's this great feedback system, measurement system.”
Many education analysts agree that traditional teacher-evaluation practices haven't been of much use. “A principal sitting in the back of the room checking off things on a list” of recommended teacher behaviors “made almost no sense,” partly because “it's bound to involve many very subjective judgments,” says Aaron Pallas, a professor of sociology and education at Columbia University in New York. “Almost everybody does well” on such evaluations, proving that the approach isn't very accurate or useful, he says.
Nevertheless, Pallas maintains, while old-style evaluations “provide almost no guidance about what to do” to improve one's teaching, new data-oriented evaluation systems don't either — at least so far.
Yet, rejecting the data approach means “sticking our heads in the sand,” says Valerie E. Lee, an education professor at the University of Michigan in Ann Arbor. “These things can be good so long as they're done right,” she says. That means including other measures besides standardized-test scores and being careful not to jump at untested teacher-evaluation approaches, she says. If developed and used judiciously, Lee says, a good system could control for individual differences in students, such as attendance and home life, over which a teacher has no influence. And that, she says, would make for fairer teacher-to-teacher comparisons than those that simply look at student test scores.
Donald B. Gratz, an education professor at Curry College in Milton, Mass., cited a bit of history in arguing that programs linking teacher merit pay and student test scores are ill-conceived. “In the mid-1800s, British schools and teachers were paid on the basis of the results of student examinations, for reasons much like” those cited by today's reformers, Gratz wrote. After about 30 years, however, “the testing bureaucracy had burgeoned, cheating and cramming flourished” and, with public opposition swelling “dramatically,” the practice “was abandoned as a failure.”
Basing pay on test scores poses another problem, too, Gratz says: Fewer than half of teachers teach subjects whose material is contained in standardized tests.
Furthermore, Gratz notes, at grades six and up, students typically have six or seven different teachers during a given year. “Who gets the credit or the blame” for a student's success or failure?” he asks. “It looks like a field day for labor lawyers.”
Offering merit pay for good teaching hasn't been shown to improve instruction either, Gratz argues. Instituting merit-pay programs “assumes that teachers know what to do and just aren't doing it,” but that's likely not the case, he says. “We do know a lot about how to teach,” but teaching is an extremely complex task, and it's not as easy as it may seem for teachers to change their behavior to incorporate research findings about student learning, for example, he says.
Complicating matters is the fact that educators' and education administrators' ability to succeed relates to the situation in which they're working, says Jeffrey Henig, a professor of education at Columbia. “We have superintendents and principals, for example, who succeed in one school, then go somewhere else and fail,” he notes. “So the question of whether someone is capable or not is way more complex than it may seem on the surface.”
— Marcia Clemmitt