By Jay Mathews | Washington Post
Monday, November 17, 2008
For a while, the fight over how to improve public schools seemed to be quieting down. During the presidential campaign, Republican and Democratic education advisers happily finished each other's sentences on such issues as expanding charter schools, recruiting better teachers and, in particular, rating schools by how much students improve.
Moving to the growth model for school assessment, by measuring each student's progress, seems to be the favorite education reform of the incoming Obama administration. Up till now, we have measured schools by comparing the average student score one year with the average for the previous year's students. It was like rating pumpkin farmers by comparing this year's crop with last year's rather than by how much growth they managed to coax out of each pumpkin.
The growth model appeals to parents because it focuses on each child. It gives researchers a clearer picture of what affects student achievement and what does not. Officials throughout the Washington area have joined the growth model (sometimes called "value-added") fan club. The next step would be to use the same data to see which teachers add the most value to their students each year.
Of course, as often happens in education, that lovely consensus is proving too good to be true, mostly because of the teacher issue. The bad news was delivered recently by Education Week reporter Stephen Sawchuk, who has been checking how the growth model was actually being received by state politicians. It turns out some legislators have been building trapdoors under the welcome mats. California banned in 2006 any use of student growth data in teacher evaluations or compensation decisions. New York last year prohibited the use of such data for tenure decisions for at least two years. Other states are staying away from anything that ties student success to a teacher's pay or job security.
Go ahead. Blame the teacher unions. They make no apology for their opposition to this approach. But they have good arguments. Congress will have to revise the No Child Left Behind law to install the growth model, and most support for the idea there extends only to rating schools, not teachers. Assessing instructors by how much their students improve seems reasonable to people like me who have never taken a psychometrics course, but nobody has sufficiently tested the statistical devices for doing that, and they might prove to be expensive.
I asked two National Education Association officials, Joel Packer, director of education policy and practice, and Bill Raabe, director of collective bargaining and member advocacy, why we couldn't test students in September and May, calculate how much they improved and use that information in deciding whether to keep particular teachers and how much to pay them. Raabe said that would only work if the distribution of students in classes was randomized. I understood his point but did not see why good teachers couldn't show some progress no matter what sort of students they have. Raabe and Packer sent me more quotes from experts who weren't any clearer.
People who have studied the public schools that have significantly raised the achievement levels of impoverished students tend to accept the idea that teachers' salaries and jobs must eventually be tied to classroom results. "Of course," said Andrew Rotherham, an education think tank founder, blogger and Virginia Board of Education member who, at 37, is likely to be a major player on this issue for years to come.
But Rotherham hastened to add that his state is going to take it slow. A poorly planned system to fire teachers based on growth data could bring lawsuits. "We are never going to get to where we need to go without more research," he said. Nancy S. Grasmick, Maryland state superintendent of schools, agreed. "The decisions we make to strengthen schools should not be made by speculation or unproven theory, but on data." President-elect Barack Obama is likely to nod when he hears that.
I asked Raabe and Packer about an alternative approach -- rating schools, not teachers, and replacing the principal if students do not improve. The NEA is not comfortable with that, either. Making the principal responsible for creating the conditions for student growth works only if the principal has the power to hire and fire staff, and that is a no-no for the association.
Could Obama broker a compromise? Randi Weingarten, president of the American Federation of Teachers, told me that she wanted teachers to learn how to use growth data "and then see where we go from there." Would the unions agree to giving principals firing power in exchange for using not just tests but also the student work portfolios and public presentations that Raabe and Packer endorsed as authentic ways to judge how well kids have been taught?
Many Republican and Democratic policy experts want to try something like that. The growth model in some form will have its day, sooner or later.
But about those pumpkin farmers: It seems to me that the best measure of their work is not how much their big, orange vegetables grow but how well they stand up to my inexpert carving for Halloween and how many extra slices of pumpkin pie I eat on Thanksgiving.
The human equivalent to that kind of quality assessment is how well each school's students do in college, in jobs and in life. I don't think we have a growth model that measures that, but people are working on it.