Do We Need National Achievement Exam? No: It Would Damage, Not Improve, Education
U.S. policymakers are besieged with proposals for a national test or examination system. The plans range from a national multiple-choice exam to a complex system of exams which are to be calibrated to one another.
While the examination proposals have significant differences, all are based on the false premise that measurement by itself will produce positive change. Recent history shows this is not true: During the 1980's, U.S. schoolchildren became probably the most over-tested students in the world-but the desired educational improvements did not occur. FairTest research indicates that our schools now give more than 200 million standardized exams each year. The typical student must take several dozen before graduating. Adding more testing will no more improve education than taking the temperature of a patient more often will reduce his or her fever.
The proposals also share the assumption that the United States needs a national exam because our education system is failing to produce workers as skilled as those produced by economic competitors such as Japan and Germany. Education in this country does need major improvements, and not just for economic reasons. But neither Germany nor Japan has a national examination system of the sort being proposed for the United States. In fact, Germany does not even have a national curriculum. If these nations provide a better education to more of their children, it cannot be because they have national tests.
In response to the national-testing proposals, FairTest and over two dozen major education, civil-rights, and advocacy groups-including the National Education Association, the National Association for the Advancement of Colored People, the National PTA, the National Association for the Education of Young Children, the American Association of School Administrators, and the national associations of elementary- and secondary-school principals released a statement urging "the Bush Administration and the Congress to support education reform by not implementing a national exam at this time." The organizations agree that mandating a national exam is premature at best and could lead to deepening educational disaster.
The scope of the potential damage is most clear in the Educate America proposal. That group seeks to administer a series of six tests to each high-school senior for $30 per student. It also claims its tests would be "state of the art" and include performance-based components. But the Scholastic Aptitude Test, which is entirely multiple-choice, costs $16 for just two tests. At the proposed price, the Educate America plan would have to be a multiple-choice test.
There can be no doubt that schools would be forced to teach to such a test. Yet organizing schooling around multiple-choice tests has been convincingly shown to do great damage to curriculum and instruction. The harm is greatest for students in the lower tracks whose schooling often is reduced to "drill and kill" to raise test scores. This method of instruction virtually guarantees they will not learn higher-order academic thinking skills.
Examination systems like those proposed by the University of Pittsburgh researcher Lauren Resnick and Marc Tucker, president of the National Center on Education and the Economy, do have positive features. Unlike Educate America, their plan calls for performance- based exams and would not be one-test-for-all. Its proponents recognize that we must develop educational standards before we implement an exam and they seem aware that assessment reform cannot be implemented without other educational changes, though the actual proposals fail to address this fact.
Indeed, assessment should be part of school reform-not the controlling force that national testing proposals make it. By focusing on assessment as the solution to our educational problems, we may well fail to address such critical issues as equity, rigid school governance, low-quality textbooks and curricula, inadequate schools of education, and a lack of useful information about school inputs, processes, and outcomes. To make real use of performance-based assessments requires creating performance-based schools, which in turn require restructuring, staff development, and new educational materials.
Any national exam system should be based on national standards. The best current example is the National Council of Teachers of Mathematics' Standards for Curriculum and Instruction. Developing these standards took several years and involved substantial grassroots participation. There is no reason to believe standards in other subjects can be developed more quickly.
Then there is the even bigger problem of developing teachers' ability to implement the new standards. Teachers do not automatically know how to teach in a restructured environment, though a great many are willing to learn. We should build on their willingness, not dump a new assessment system on them, fail to provide adequate support, then penalize them and their students for not doing well.
FairTest is also concerned that these examinations could become a national gatekeeper that continues our nation's unfortunate history of unfairly sorting students by race and class. Barring additional changes, it is all too likely that districts will sort students according to their perceptions of how rapidly students will advance toward the "certificate of initial mastery" proposed by Lauren Resnick and Marc Tucker. Such sorting will not spur low-income and minority-group students to improved achievement.
On any exam, some students who fail should have passed. Experience shows that students from low-income and minority-group backgrounds will be disproportionately those who will suffer the negative consequences of false failures.
Exam proponents often say that they don't want to unfairly penalize students and that they support "second chance" systems. But why should we believe that "second chance” programs will be adequately implemented? We know WIC and Head Start work, but our nation does not fund them adequately.
Performance-based assessments can be made reliable, sensitive to potential bias, and even helpful in addressing issues of bias. But work to assure these assessments are bias-free is only just beginning.
The proposals also leave key questions unanswered: Who will set the standards, develop the exams, and establish the scoring criteria? Imposing a national exam could lead, without adequate public discussion, to a national curriculum and an unelected de facto national school board that will erode democratic control of education and local accountability.
We do not even know whether it is feasible to construct such a national system. The whole process, particularly the calibration of possibly hundreds of different exams to each other, could prove to be too expensive and unwieldy to work. When the complexities become clear, the portfolios and projects necessary for performance-based assessment could end up being reduced to very limited exams. There could even be a return to multiple-choice tests, with all their well-known flaws.
Performance-based assessment methods can assess higher-order abilities and encourage good educational practices. However, we can move toward the national use of such assessments without constructing a national examination system. We then gain the advantages of good assessment and avoid the dangers of imposing a national testing system.
Assessment reform should be incorporated into systemic educational reconstruction at all levels. We must begin by defining the kind of education we want our children to have, including both their daily experiences and the outcomes society desires. On that basis, we can determine how to make the changes in curriculum, instruction, school governance and structure, and assessment required to reach educational goals far more comprehensive than those enunciated by the Bush Administration and the governors.
To do that, the pieces of a reform program must be organized into a coherent whole. Only after we have real experience in implementing the changes will we have the information necessary to make a reasoned decision about a national test. Once these reforms have taken hold in classrooms, schools, districts, and states, there may be no real need for the expense and complexity of a national exam system.
In the interim, we do need changes in assessment, as we need reform in all areas of education. For one, states and districts should stop the incessant, numbing, destructive multiple-choice testing most of them now engage in. They should develop and implement performance-based assessments, but do so while changing curricula, instructional methods and materials, ensuring the staff development of teachers and administrators required to make it work, and involving parents and other members of the community in the process.
The federal government should support improvement efforts that include assessment reform and that build consensus and change from the bottom up, with guidance-not dictates- from national organizations such as the National Council of Teachers of Mathematics or bodies such as the National Goals Panel. Support for systemic educational reform would be a far better use of limited federal resources than imposing a national test or examination system would be.
Education can be dramatically improved over the next decade. Assessment reform is part of the way to make the changes, but it is not the magic key. Just testing without ensuring all the other necessary changes is a prescription for failure, a false short-cut that will actually undermine education reform. Public education in the United States can ill afford such an error.
Vol. 10, Issue 31, Pages 28, 36Published in Print: April 24, 1991, as Do We Need National Achievement Exam? No: It Would Damage, Not Improve, Education