The Perils and Possibilities of Teacher Evaluation
In the film Waiting for Superman, Harlem Children’s Zone founder Geoffrey Canada talks about his experiences as a beginning teacher in Boston. After completing a master’s degree at Harvard’s Graduate School of Education, Canada taught at a private alternative high school for troubled youth. His naïve bravado—expecting that it might take him a year or two to “straighten out” education in the U.S.—invariably draws an appreciative chuckle from the audience, as does his admission that he was a terrible teacher during his first few years of teaching.
Canada is so charismatic, and obviously cares so deeply about the children in his charge, that it seems ungracious to point out that we have only his word that he became a good teacher. It’s not hard to imagine that even someone as intelligent and motivated as Canada might have struggled at the outset; the term “rookie mistake” is commonly used to describe the learning curve of novices in many occupations. But what evidence do we have that he became a skilled teacher?
Until recently, judgments about whether a teacher is “good” or “bad” have rested on vague criteria and evidence. In many school districts, nearly all of the teachers who have chosen to stick it out for three years or more have been rated satisfactory, which is the key step in achieving the due process protections of tenure. Some critics wonder how the vast majority of teachers can be rated wholly satisfactory when many students are not meeting their state’s standards for academic proficiency.
The answer lies in our uncertainty about how to measure good teaching. Clearly, it’s inadequate to judge good teaching based on what a principal might see upon dropping by a classroom for a few minutes a few times a year. But how are we to judge who is a good teacher? Currently, there are two main thrusts. The first focuses on the idea that a good teacher is one who contributes to her students’ learning.
This idea is more challenging than it sounds, for two reasons. First, Americans expect and demand a great deal from our schools and teachers; and although the learning of school subjects such as reading and mathematics is a central goal, it’s not the only one that we think is important. New York City Schools Chancellor Joel Klein, commenting on his most memorable teacher, science teacher Sidney Harris at Bryant High School, once said, “Mr. Harris’s expertise was in physics but what he taught me went far beyond science. He pushed me. He shaped the way I thought about my future. And he set expectations for me that were, before then, unimaginable.” Chancellor Klein, like many of us, valued the way in which a teacher might incite a sense of curiosity about a subject, instill a sense of purpose and direction to one’s life, and spark creativity and commitment. There’s little reason to think that these important types of learning would be picked up by a standardized test of mastery of physics, or of any other school subject. Without doubt, it might be difficult to determine how to assess how teachers contribute to these kinds of longer-term outcomes; but that’s not a reason to ignore them.
Second, the notion that a teacher contributes to her students’ learning of school subjects implies that we have some sense of what those students knew before they came into contact with that teacher. In recent years, a statistical technique called value-added measurement has sought to isolate the impact of a particular teacher on student achievement, using the standardized tests which students in the elementary and middle school grades take annually. But such measures, which are the basis for the controversial Teacher Data Reports in New York City, are an imprecise method for identifying good teachers, and in any event are only applicable to teachers who teach subjects which are tested each year. Moreover, they tell us nothing about the nature of good teaching. Experts agree that value-added measures of teacher effectiveness should not be used as the primary basis for evaluating teachers, especially for high-stakes decisions like tenure.
The second approach to identifying good teachers is focusing on their practices. Teaching is an extraordinarily complex activity, and because classrooms and students differ in so many unpredictable ways, no two instances of teaching are ever identical. There is, nevertheless, a growing consensus about the teaching practices associated with planning, instruction and assessment which maximize the likelihood that students will learn. Observing and rating classroom practice is thus another strategy for identifying good teachers. With proper training, school principals and teachers can observe classroom teaching and agree on what they see and what it means.
Both approaches to identifying good teachers are in their infancy, although there is good reason to think that we can get better at both of them. Neither approach is likely to be satisfactory in isolation from the other. If we are willing to invest in the development of better measures of student learning and of good teaching practice, we can do much better than simply asserting, with a minimum of evidence, that a particular teacher is either good or bad at his job. #
Aaron Pallas, Ph.D. is Professor of Sociology and Education at Teachers College, Columbia University.