“I know you’re restless today, but I need to see you sitting at your desks. Angel, that means you, too!” In the second-grade classroom at the Washington school where I volunteer, the teacher turned to me and said with a sigh, “It’s testing week.” In fact, her class wasn’t suffering through the standardized ordeal, just tiptoeing around while others did. The “adequate yearly progress” (A.Y.P.) assessments mandated by the No Child Left Behind legislation, which was enacted in 2002 with high hopes of closing the achievement gap for minorities, don’t kick in until third grade. But when it comes to tests, N.C.L.B. is fulfilling its inclusive mission all too well: nobody — not even kids too young to be filling in the bubbles yet — escapes the atmosphere of exam-induced edginess.
The president’s signature domestic initiative, now due for its five-year reauthorization, was supposed to be a model of the hardheaded rigor it aims to instill in America’s schools. “No ‘accountability proposals’ without accountability,” a Bush education adviser declared early on. So one of the most glaring legacies of No Child Left Behind is surprising: it has made a muddle of meaningful assessment. Testing has never been more important; inadequate annual progress toward “proficiency” triggers sanctions on schools. Yet testing has never been more suspect, either. The very zeal for accountability is confusing the quest for consistent academic expectations across the country.
In 2014, when states are supposed to report 100 percent pass rates, no governor will be able (honestly) to claim perfect success. But by then, it would be useful at least to agree on what “proficiency” entails. That issue is precisely what is obscured by a blizzard of scores, courtesy of America’s decentralized educational tradition. N.C.L.B. left the states free to choose their own standards and testing methods for determining adequate yearly progress toward proficiency in math and reading. The data therefore defy comparison. In Florida, for example, 71 percent of schools failed to make A.Y.P. in 2006, while only 4 percent did in Wisconsin. More brain-boosting cheese on school lunch menus, perhaps?
Problems don’t end there — just as a social-science principle called Campbell’s law would predict. “The more any quantitative social indicator is used for social decision making,” the social psychologist Donald Campbell concluded in 1975, “the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.” With “high stakes” testing, N.C.L.B. introduces an incentive not to cheat, necessarily, but to manipulate. Signs are that states define proficiency down while schools ramp up narrow test prep. “Score pollution” — results that reflect intensive coaching — becomes a risk.
And all for what? Not leaps in learning, to judge by an older, federally financed test called the National Assessment of Educational Progress, whose format reflects clear standards for basic subjects and goes beyond multiple-choice questions. Developed by a nonpolitical group of educators, subject-matter specialists and nonexperts, it is administered every two years to a representative sample of students in grades 4, 8 and 12. The purpose of the test, known as “the nation’s report card,” is diagnostic, no strings attached. Its results are sobering. While the states’ tests typically show rising math and reading scores, with roughly 70 percent of students rated proficient or better, the National Assessment reports only about half that proportion scoring so well.
Angel, happy to escape his seat and read a book with me (he chose “Flat Stanley”), would doubtless be thrilled if test week disappeared. It won’t. But the test mess could be what is called in the trade a teachable moment, a chance to consider the case for national standards and a single national exam. There’s nothing like a blend of confusion and coercion at the state level to make the prospect of credible countrywide assessments — based on coherent expectations of what students should know — look less like creeping federal intrusion and more like welcome clarity.
Where ideological clashes doomed a quest for national standards in the 1990s, pragmatic calculations might tip the balance in favor now. Let the federal government pay for a national test and the formulation of standards, suggests Diane Ravitch, a clear-eyed historian of school reforms who was an assistant secretary of education during George H.W. Bush’s administration. The move could save the states the estimated half-billion dollars they spend on their own testing programs — and, Ravitch notes, give the U.S. government a job it is good at: gathering and spreading information about how states are faring.
The National Assessment of Educational Progress could serve as a model for a test that judges students’ ability to apply their knowledge and thus discourages rote coaching. But recent experience — and Campbell’s law — argues against making test results the sole trigger of federal sanctions. Instead, the data would give states and school districts reliable information on where progress is, and isn’t, happening across the country, to catalyze their own strategies to boost achievement. Rather than cramming to reach an unrealistic target by 2014, states could be more like the laboratories of curricular improvement the country needs. Agreeing on common goals for what kids should be learning can free up teachers to focus more productively on how they could be learning better.
School transformation can’t be engineered by any test, which is a two-dimensional tool at best. Still, a good national exam would spread well-focused standards across state borders and spur progress. Reading Angel’s book to him, I saw an apt metaphor: poor Stanley wakes up to find himself flattened, a boy become a board, but the discovery that he can slip into an envelope and travel around the country expands his horizons.
