A Model Idea

NCLB currently uses a "status model" in its school and district ratings and focuses on the percentage of students that meet state standards on standardized tests in a given year. Growth models (often called value-added models), by contrast, measure changes in student performance over time. If well-implemented, the new ED guidelines could allow states to more fairly and accurately identify schools that make significant progress with under-performing students, even if those students still fall short of the proficient goal. And they would give educators an incentive to raise the achievement of all students, not just those close to passing state standards.

Badly implemented growth models, however, have the potential to water down standards where they're needed most. In that respect, the barely-restrained enthusiasm of the avowed NCLB critics at the National Education Association—whose anti-NCLB lawsuit was recently thrown out of court—was telling. NEA President Reg Weaver said, "Today the Department heeded to the calls of millions of educators for a 'growth model' that truly reflects the great progress we are making in the classroom."

Note the tense: "are making." If NCLB is altered to validate current performance levels, it will no longer provide external pressure to enact the kind of fundamental education reforms generally opposed by status quo defenders like the NEA. This fear was echoed in concerns voiced by NCLB supporters such as the Citizens Commission for Civil Rights, the Education Trust, and key lawmakers like Senator Edward Kennedy (D-MA) and Representative George Miller (D-CA), the ranking Democrats on the Senate and House committees on education.

The guidelines themselves are complex, reflecting the intricacies of measuring growth for multiple subjects, grades, student-subgroups, and levels of school governance. In the long run, this policy will lead toward greater inconsistency from state to state. ED correctly notes that "no two models will likely be alike." But that will take some time -- tracking growth in student learning means tracking students from grade to grade and school to school along a series of aligned assessments, which is a feat beyond the technical capacity of most states.

But the single most important issue relating to growth models was only partly addressed: On what basis do states decide how much growth is enough growth? The guidelines stipulate that a "model that only expects 'one year of progress for one year of instruction' will not suffice, as it would not be rigorous enough to close the achievement gap by 2014." But it is left to the states to determine how much more than a year of growth is rigorous enough.

This issue is crucial. NCLB was designed to impose an absolute standard on public school performance: students have learned what they need to learn, or they have not. It makes no allowances for extenuating factors or plain excuses. Student proficiency matters most. In the law, it even trumps the fact that some schools face far greater challenges than other schools in getting their students up to state standards.

Growth measures, by contrast, are relative by definition—one score compared to another. Therefore, the only objective way to determine sufficient growth is to extend the trajectory of a student's growth forward in time and see where it leads. If the current rate will get them to proficiency—as measured by legitimate, rigorous standards, not just low-level high school exit exams—then it can provide a reasonable, perhaps even superior, gauge of school success. Growth models are potentially fairer and more accurate measures of school performance. But without that crucial endpoint in proficiency, however, schools will become untethered from real-world standards and inevitably drift back to the comfortable imprecision of relativity that has served students poorly for too long.

In the end, the burden will fall on the Department of Education to keep this central principle in mind. If it adheres strictly to the guidelines it has created, and insists on growth models that put students on a path to real proficiency, then this experiment could result in the best of both worlds—clear expectations for student achievement, and a more credible system of measuring school performance. If it does not—the Department's track record on NCLB enforcement is mediocre at best—then it could weaken the law in ways that make future progress even more difficult to achieve.