|
WHEN STANDARDIZED TESTS GO WRONG
By JOHN A. CHAMBERS (NYT) 'MATH Test Ruled a Failure,'' reported one Westchester paper last week; it was typical of late June headlines across the state, with the news of the Math A Regents examination results being set aside by the New York State Commissioner of Education, who said: ''This exam doesn't seem fair. I think we made some mistakes.'' Teachers, students and parents seem in overwhelming agreement, so the commissioner's decision was met with some relief, especially in cases where graduation was at stake. But that is not all -- the Physics Regents exam, for the second consecutive year, has also sparked controversy. Meanwhile, a panel of experts found that essay questions on the Regents Exam in Global Studies were ''vapid'' and ''hopelessly broad'' and noted the prevalent critique from teachers that the exam forces a curriculum ''a mile wide and an inch deep.'' From where I sit as a public school superintendent, these concerns are well placed, but they are only symptoms of still deeper problems with the current overemphasis in state and federal policy on standardized testing as a method to judge student performance. No, I am not just another superintendent defending the status quo, opposed to change and accountability; on the contrary, my time in public schools keeps teaching me how much improvement is needed, how important it is to set higher standards. I work in a school district where students do very well on nearly every standardized test, and where teachers and principals keep helping them to do better still. I also serve as a trustee of a nonprofit organization that serves the testing needs of some 1,500 schools in 41 states and 32 countries. Yet I see at least three big problems in the direction the state and nation are headed with standardized tests. Problem No. 1: Too many tests, too much emphasis. New York is one of the most test-driven states in the country, requiring five Regents exams for graduation, five tests for eighth graders, another three for fourth grade and one for fifth grade. And now the federal No Child Left Behind Act is piling on requirements for two or three tests at every grade from three to eight. Nor does it stop there -- in New York, as recent headlines indicate, the State Education Department has had great difficulties getting the tests successfully matched to the standards, and that is only going to get harder as the tests multiply over time. So much testing leaves less time for teaching, and even while that precious time gets pushed more toward test preparation, the level of the playing field keeps changing because of the many problems with the quality of the tests. Remember the last round of headlines, over altered quotations on the English Regents? Have we not seen enough? Problem No. 2: Too little flexibility, too little support. All these required tests have displaced and discouraged other approaches that work, like the highly successful portfolio assessment approach used by the Urban Academy in Manhattan and its 31 consortium partners. But those schools and others showing success with innovative methods have been denied waivers by the state's Commissioner of Education. Worse still, New York, like much of the nation, still suffers from a two-tier education system, where affluent suburban schools usually have higher test scores, and inner city schools, where children may not even have textbooks, typically have lower test scores. If high-stakes testing was supposed to change that, we have seen little indication of it so far in New York, where the arcane state aid formula keeps the existing inequities intact. Problem No. 3: Research suggests that all this rigorous testing is of dubious value. If we could solve the first two problems, then would high stakes testing be a good idea? No. The largest study ever done on such testing, performed by researchers at Arizona State University in 2002, found that make-or-break testing that decides whether students graduate, ''an approach already in place in more than half the nation, does little to improve achievement and may actually worsen academic performance and dropout rates,'' according to a New York Times article on the study. The study, according to the article, ''found that while students show consistent improvement on these state exams, the opposite is typically true of their performance on other, independent measures of academic achievement.'' So what to do? Better policy, while less reliant on high stakes tests, could still make good use of data from fewer, more carefully designed standardized tests. But it would also include other measures of performance in more authentic settings. In school as in the workplace, a paper and pencil test alone can never fully measure ability or achievement. We should be encouraging innovation rather than a one-size-fits-all approach, recognizing what the overwhelming majority of people who work in America's schools, and the students who take all those tests in those schools, already know: Every student is different, complex and deserving of more than an endless battery of tests. But because better policy will also be more complex, more time consuming, and less easily expressed in a political sound bite or a neat table of test scores, it will not come about until the people of New York, and the nation, insist on what children deserve. And they deserve resources where the need is greatest -- not least of all to close the achievement gap which is one of America's most shameful continuing scandals. |
