Sent to the Seattle Times, August 30. 2010
Dan Goldhaber ("Getting ahead of the teacher-accountability curve," August 29) points out some of the problems with using growth on standardized tests (value-added analyses) to measure teacher effectiveness. He points out that small differences may not be statistically significant and that factors other than the teacher influence test scores. Goldhaber, however, still supports using value-added methods "as one means of judging teacher performance."
They shouldn't be used at all. Studies show that value-added scores are unstable and inaccurate. The same teacher can get very different scores from one year to the next, and different reading tests produce different value-added results. Also, test scores can be artifically pumped up by teaching test-taking strategies.
I am opposed to using student test scores as part of teacher evaluations, not because I am opposed to evaluation but because they are poor measures of teacher quality.
Getting ahead of the teacher-accountability curve
Seattle Times, August 29
Among the issues in contract negotiations between Seattle Public Schools and its teachers union is a plan to include improvements in student growth as part of teacher evaluations. Guest columnist Dan Goldhaber supports the idea of using value-added methods as one means of judging teacher performance, but notes the importance of careful planning and implementation.
By Dan Goldhaber
SEATTLE Public Schools and the Seattle Education Association are currently engaged in high-stakes negotiations over a new contract.
A key sticking point in the negotiations appears to be whether student test scores ought to factor in to teacher evaluations, employment or compensation.
There are powerful reasons for thinking outside what has been a very narrow box in terms of teacher compensation. Teacher-policy systems in this country typically do not recognize or act upon the significant differences we know exist between teachers, nor do they recognize the different circumstances in which they teach. Teachers serve in very different schools and have different labor-market opportunities outside the classroom. And yes, they contribute differently to student learning in ways that show up in research studies and are obvious to parents and fellow teachers.
The pressure is on to get outside the narrow box now. Like it or not, we must face the fact that the horse has left the barn when it comes to teacher accountability.
This is perhaps best exemplified by a story that recently ran in the Los Angeles Times. The story focused on the effectiveness of teachers serving in the Los Angeles Unified School District and took the district to task for ignoring the huge disparities that exist in teacher effectiveness. Using the district's own data, the paper commissioned a value-added study (a statistical methodology designed to isolate the contributions that teachers make toward student achievement from other factors that we know affect student learning).
Moreover, the LA Times has promised in the future to publish the names of individual teachers and their value-added scores. This promised release of data would make the estimates of teacher performance very public, an idea that was just endorsed by U.S. Education Secretary Arne Duncan.
I support the idea of using value-added methods as one means of judging teacher performance, but strongly oppose making the performance estimates of individual teachers public in this way.
• First, there are reasons to be concerned that individual value-added estimates may be misleading indicators of true teacher performance. Teachers may not, for example, be fully responsible for the learning of all the students in their classes. Pullout programs, migration of new students into classrooms, and other ways in which instruction is specialized in schools, make the attribution of students to teachers complex.
• Second, performance estimates that look different from one another on paper may not truly be distinct in a statistically significant sense. Addressing details like these is not an insurmountable hurdle, but it requires a commitment to building a performance-evaluation infrastructure.
• Finally, and perhaps most important, I cannot think of a profession in either the public or private sector where individual employee performance estimates are made public in a newspaper.
I do know there is frustration over the collective failure to act on differences in teacher performance. Unless there is greater willingness to experiment with reforms, I fear debates over teacher policies will grow increasingly contentious.
The situation that is beginning to be played out in Los Angeles shows the pitfalls of inaction that come out of polarization and political constraints, leaving school systems institutionally incapable of differentiating among teachers.
If schools and teachers unions do not get out in front of the teacher-accountability curve, they both may get run over.
We are at a crucial time in the teacher-accountability movement. Change, however, is not easy. Successful reform requires careful planning and implementation, and a willingness to engage in evaluation that leads to midcourse corrections.
Steady progress requires the education enterprise to become a continuously evolving system where it is common practice to learn from the policies that are enacted and make adjustments to them based on what is learned.
Dan Goldhaber is the director of the Center for Education Data and Research at the University of Washington Bothell.