Voices in Education

Not by “Value-Added” Alone
Publicly reporting test scores for entire schools is one of the more positive and logical educational innovations in recent years. Perhaps this is why someone at the L.A. Times thought it might be a good idea to take this one step further and report scores for individual teachers. Or perhaps someone just wanted to make headlines. Did they succeed?

Headlines? Yes. Good idea? No.

To be fair, consider the potential usefulness of this approach: Teacher performance varies widely even within individual schools, and performance is not very tightly linked to teacher credentials like certification and degrees. Further, expanded testing and improved data systems has made it increasingly possible to estimate what teachers contribute to student learning—what researchers like me call “value-added.” But judging teachers by this measure alone ignores three major problems:
  1. We can only estimate teacher value-added. We do not know for sure what these teachers really contribute to student learning. Like any statistical inferences, value-added measures entail a range of error. If the L.A. Times decides to do this report again next year, most of the best-rated teachers will no longer be on the list—not because the teachers got worse, but because they may not have belonged on the list to begin with.
     
  2. Even if we could determine teacher value-added with certainty, these measures only capture teachers’ contribution to student test scores. No state (certainly not California) has a test that is so good that we could rely on the test alone. Even if we do eventually develop and use tests that capture higher-order thinking and writing skills, some things we expect from schools cannot be tested this way—creativity, curiosity, working in groups, and love of learning, to name a few.
     
  3. Even if we eventually create perfect measures of teacher effectiveness, releasing them publicly will only wreak havoc on schools and undermine teaching and learning. Sure, the barrage of phone calls and lobbying from parents will put pressure on schools to improve, but there are more constructive ways to accomplish this at the school level. And Lake Wobegon this is not—no matter how much schools improve, some teachers will always be below average. Are we really going to make a policy of having half of parents run for the doors when they see that their children have not been assigned to the top teachers? Private schools don’t do this, nor do private businesses, so why force a bad idea on public schools?
Value-added measures of school and teacher performance have many potentially positive uses, especially when combined with other measures that capture a broader range of educational goals and important instructional practices. But the L.A. Times—and the Los Angeles school district, which facilitated the story—have taken the wrong approach.

Let’s hope that other districts avoid making the same mistake. We are in an unprecedented era of educational reform, most of it positive. In contrast, the L.A. Times’ ill-conceived, headline-grabbing approach will only breed mistrust among the teachers being evaluated, without doing anything to actually help students.

About the Author: Douglas N. Harris is associate professor of educational policy and public affairs at the University of Wisconsin at Madison. He is author of a book on value-added measurement and school performance, forthcoming from Harvard Education Press in spring 2011.