® Clearinghouse on Tests, Measurement, and Evaluation April 1992 EDO-TM-92-1
® Clearinghouse on Tests, Measurement, and Evaluation April 1992 EDO-TM-92-1
Lynn Davey, ERIC/TM
In the last decade, our dissatisfaction with the quality of American education has grown. Perhaps A Nation at Risk (1983) most clearly voiced that dissatisfaction. Experts have proposed an array of remedies, many focused on making schools accountable for what students learn. If we specify the goals we want to attain and hold teachers and administrators responsible, we've assumed, students would learn more.
Unfortunately, the tests used to fuel the fires of accountability have shed a lot of heat but little light. The measures--ranging from the Scholastic Aptitude Test to the National Assessment of Educational Progress to standardized norm-referenced tests used in state and local assessments--have proved inadequate. They have failed to provide the information we need about students and specific curricular objectives. So far, they have not helped us improve our schools.
As currently formulated, these tests often
assess only a narrow range of the curriculum;
focus on aptitude, not specific curriculum objectives; and
yield results that apply only to the state or nation, not the individual pupil and school.
After the years of bringing us good news during an era of declining performance, today these tests have little credibility.
A new national assessment system, however, could lay a solid foundation for improving schools. Such a system would allow us to monitor schools' and teachers' effectiveness, and students' progress toward national educational goals. While serving accountability efforts, the system would also strengthen and expand students' intellectual scope.
A sound national assessment system would
allow valid comparisons between schools, districts, and states;
provide sound data on achievement, not just aptitude
prod schools to focus on important, performance-based outcomes; and
yield results for every important level of the education system, from individual children to the nation as a whole.
A carefully designed and implemented national examination system can fulfill these needs--
without requiring a national curriculum, and
while still increasing incentives for all students to achieve.
We need a better system for comparing schools, districts, and states
Because the United States has no national system of achievement testing, we cannot validly compare students' performance across the nation. Parents, educators, and policy makers who want to know whether a child is learning as much as others, or whether a school or district is effective, must base their analysis on limited data of dubious quality.
The scores reported for many widely used achievement tests are, in fact, more likely to mislead parents and the public than to enlighten them. Instead of showing how students perform compared with real standards, these exams rate test takers against samples of students who took the tests before--sometimes many years before. And so, most states are reporting that their students, like the children in Garrison Keillor's mythic Lake Wobegon, perform above average (Cannell, 1987).
We need to know how our children and schools are doing. Nothing we have so far, including the National Assessment of Educational Progress, permits valid comparisons. To create effective accountability systems, we first need useful and reliable data on student outcomes. Nationwide tests for all students could give parents, teachers, and policy makers that data.
We need sound data on achievement, not just aptitude
The SAT and ACT are the closest things we have to a national achievement test. The SAT, in particular, affects many college-bound students. However, the students who take the test represent a subset of all high school graduates, and the SAT focuses on aptitude, not achievement. It does not assess what students have learned about subjects like science, history, or geography.
Thus, our most common, high-stakes tests are estranged from the curriculum and fail to reinforce the notion that hard work in school matters. In fact, students who study hard and learn a lot receive few concrete rewards. Little discomfort befalls students who just slide by. By contrast, the achievement tests given in other countries convey the idea that mastering school subjects is important. Those exams make students accountable for what they have learned.
A national achievement test will force schools to focus on important, performance-based outcomes
Some proposals for a national achievement test include writing samples. Others would require students to give practical demonstrations in subjects like music and the natural sciences. Most advocates of national examinations stress the value of performance-based assessments. They argue that students should demonstrate whether they can organize their thoughts, analyze information, and formulate arguments. In short, tests need to determine whether students can apply the knowledge they have learned.
Scoring performance-based exams is more complicated than scoring multiple-choice exams. But the experience of other countries like Germany, England, and Japan shows that this type of exam works. With proper training and monitoring, different people can learn to assess performance-based exams according to a common standard.
A national achievement test does not have to mean a national curriculum
Some countries with national exams--France and Japan, for instance--have strict national curriculums. Germany, like the United States, considers education mainly the responsibility of the individual states. Each state follows its own curriculum; though all states follow the same format for the German exam, each decides on the specific questions it will ask.
Achievement tests work best when they explicitly relate to curriculums. The German experience shows that even when curriculum and assessment fall under local authority, a national examination system can still work.
If we hold all U.S. schools to a shared standard, however, we will need to coordinate our efforts. Just as a national system of achievement testing need not entail a national curriculum, it need not entail a single test for everyone. We can construct multiple tests for different subjects.
A national achievement test could increase incentives for all students
Unlike the SAT, the ACT, and the national exams of most other countries, the achievement tests now being proposed in the United States are aimed at all students. If employers started asking for transcripts and stressing academic achievement when hiring, scores on such tests could become a valuable credential for students entering college as well as the workplace. The labor market, along with the school, needs to reward academic achievement. The high expectations that achievement tests raise and the incentives they provide should be part of all students' education, regardless of their aspirations.
A testing system needs to respond to differences in people's educational and career aspirations, in their intellectual acuity and commitment, and in their cultural values and religious beliefs. But we also need to focus on one main goal: to see the skills and knowledge of the average student emerging from the average school clearly rise.
Conclusions
Achievement tests should measure what we think is most important for students to learn. However, our curriculum objectives, pedagogy, and testing programs have seldom been aligned.
New tests must provide accurate, reliable information:
Tests are our primary source of information about educational achievement;
To improve American education today, we need to hold schools accountable for students' performance.
To hold schools accountable, we need reliable information about student outcomes.
Any proposed national examination must be part of a broader plan, a plan that also integrates objectives, standards, teaching, assessment, and accountability for results.
References
Bishop, John H. (1989). Why the apathy in American high schools? Educational Researcher.
Cannell, John Jacob. (1987). Nationally normed educational achievement testing in America's public schools: How all fifty states are above the national average. Daniels, WV: Friends for Education.
Finn, Chester, Jr. (1991). We must take charge: Our schools and our future. New York: Free Press.
U.S. Department of Education (1983). A Nation at Risk. Washington, DC: Author.
Sitemap 1 - Sitemap 2 - Sitemap 3 - Sitemap 4 - Sitemap 5 - Sitemap 6