This appearance of federal magnanimity was driven by the recent discovery by the federal government that eventually, every school will fail. Of course, this has been a well-known fact for years. Thus, “waiving” the requirement that all students be proficient by 2014 is a bit like waiving the prohibition on getting wet after the tsunami has hit the beach.
The “flexibility” word is used as the lead in ten straight paragraphs of the waiver provisions. Unfortunately, this is an exercise in Orwellian Newspeak. For those thinking the waivers offer new freedom and flexibility, they need only review the federal twenty page list of questions and conditions set forth in the review guidance. States and locals are free to be flexible as long as they adopt a national curriculum for all students, apply one of two sets of common national assessments in grades 3-8 plus high schools, adopt a set of supports/ sanctions for the lowest scoring 15% of schools in each state, embrace at least three student performance levels, and apply growth models to evaluate teachers and principals by standardized test scores.
The waiver model is basically a rebranded NCLB with cosmetic changes. At the core, it is still a top-down, test-based, accountability model with the very same required reading and math tests. The “flexibility” is that states can add even more tests to a system that needs relief from over-testing. Amazingly, these notions are proposed at the same time that the prestigious National Academies reportshows that test based accountability systems have shown no evidence of working and, in fact, cause drop-out rates to increase. After a decade’s experience with NCLB plus almost another decade of Goals 2000, Cuban reports that high stakes accountability has failed to register a single major success in an urban high school district.
Increasing federal mandates by coloring them as freedom and flexibility is probably most troubling as relates to the mandated growth scores. Most everyone intuitively embraces growth scores. They make sense on their face – compare last year’s scores with this year’s scores and see how much progress has been made. It’s the latest panacea and the federal government wants to apply this technique to evaluate teachers, principals, schools, districts and even teacher preparation colleges.
As appealing as this concept, there simply is no acceptable way of ethically using growth scores in “high-stakes” applications. The problem is straight-forward. In order to measure growth, the pre-test and post-test must be “vertically equated” so a comparable difference can be calculated. There’s no good way of doing this. To work, vertical scaling assumes that learning progresses in a straight line from easy to hard. The problem is that children don’t learn in quite so orderly a way! (No parent who has had two or more children questions this fact). Then, complex formulas are used to account for things such as poverty, starting points, and the like. The problem is that the measures used in these models are quite sloppy and imprecise. The result is highly erroneous scores.
For example, when Sean Corcoran examined New York City teacher growth scores, he found an error band that reached 60 percentile points. A re-analysis of the Los Angeles Times teacher test data with a slightly different yet sensible model shows that the proficiency category changed for 54% of the teachers. Growth models are useful for gross comparisons but when high stakes for students, teachers or schools are attached to these results, the consequences can be destructive. Yet, the NCLB waivers require such erroneous approaches be pursued. The misapplication of these models is more than a technical question; it is an ethical and moral question.
Of course, all of this assumes that the purpose of education is increased test scores. Most all teachers, parents and citizens reject such a narrow focus. The nation’s curriculum has already been narrowed through NCLB and despite claims of “higher-order twenty-first century skills,” the standardization of curriculum to only those things that are uniformly measurable in a statistically reliable way thwarts the very skills the nation and our children need.
The problem of too many schools being labeled as failures by NCLB is simply erased by the convenience of labeling only the lowest scoring 15% of schools in each state as failures. It takes no particular insight to know that these schools will be the ones with the least affluent students and the greatest proportion of children of color. On average, these schools spend less and provide fewer opportunities. While some claim that NCLB “focused the spotlight” on needy schools, this did not translate into tangible support or assistance to our most needy.
With the federal budget in crisis and states still in recession, the new costs for statewide data bases, assistance to needy schools, and the hardware and software required for the new computerized standardized tests remains an open and unanswered question. Also, with a declining federal funding commitment, the cost-effectiveness tipping point requires that all states do their sums.
Thus, when we step back and look at the offer of waivers, we must carefully balance the promised freedom against an entirely new set of federal mandates which have massive cost, educational and ethical implications. Now that the Hosanna’s have been raised about the great virtues of waivers, instead of a stampede of applications, a more sober and extended appraisal needs to begin.