National School Testing Faces Many Roadblocks : Education: Some experts say exams have ideological, political and technical problems to overcome.

Sometime by the end of the decade, many, if not most, fourth-, eighth- and 12th-grade students in a small cluster of states will take a sophisticated series of new examinations to assess their knowledge in English, mathematics, science, history and geography.

Different sets of exams--developed separately but covering the same basic subjects and grade levels--will be administered to students in various regions across the country. Most important, all of these regionally devised tests will be graded according to a single standard, giving this country its first stab at a national exam system, long a feature of nearly every other industrialized nation.

At least that is the scenario suggested in the broad outlines of the Bush Administration’s call for a system of American Achievement Tests. Such a system is a prominent feature of the wide-ranging proposals for overhauling American education that the President and his new education secretary, former Tennessee Gov. Lamar Alexander, unveiled last month.

Although no formal plans are in place, officials hope to have at least one portion of the exams--mathematics--ready in 1994.

But there are formidable ideological, political and technical obstacles to the testing system’s development. Nothing quite like what is envisioned exists anywhere today, and there is broad rethinking, here and abroad, about how the practice of assessing students ought to be conducted.

“What the policy-makers want may outstrip people’s ability to provide it,” said Eva L. Baker, co-director of the Center for Research on Evaluation, Standards and Student Testing, a UCLA-based, national research consortium with a $15-million federal grant to find better ways of measuring student performance.

Those who remain opposed--or at least skeptical--voice concerns that the new generation of tests will share some of the problems of current assessment practices. They could unfairly categorize students, do little to improve content and methods of instruction and divert attention and money from more important educational reforms.

The call for several voluntarily taken tests instead of a single, federally required exam has helped melt much of the broad, longstanding opposition that has kept the United States from adopting a national achievement exam such as the ones administered to students in Japan, France, England, Germany and elsewhere. In most countries, all participating students take the same high-stakes exam, which in many cases is used to decide who will receive the best educations and, consequently, attain the highest stations in life.

“Having one test--instead of a system of tests that are based on standards that have been arrived at in a participatory, consensual way--is a political problem,” said Colorado Gov. Roy Romer, a Democrat who heads the National Education Goals Panel, one of several groups that helped influence the President’s education strategies. “One test is proscriptive from the national level to the local level,” a concept that flies in the face of this country’s tradition of allowing states, along with local school districts, to make most decisions about their schools.

But the so-called cluster system envisioned here--multiple tests tied to a single standard--comes with considerable technical problems: No one knows yet whether it is possible to compare disparate efforts accurately or meaningfully.

“It is widely viewed that we have taken on a big technical issue,” said Lauren Resnick, director of the Learning Research and Development Center at the University of Pittsburgh. Resnick, along with Marc Tucker of the National Center on Education and the Economy in Rochester has been working since last fall on a privately funded project to develop a pilot for a system much like that envisioned in the President’s education strategies.

“We have the right educational and social policy approach. I do agree that we now have a technical issue in (trying to develop) multiple exams calibrated to a single high standard,” Resnick said.

Some approaches being considered are embedding a core of common questions into each of the regional exams and developing “anchor” tests that would allow the results of each regional exam to be measured against a national standard.

Those who helped shape the Administration’s education reform agenda--including the goals panel and the President’s Education Policy Advisory Committee, chaired by Paul O’Neill, ALCOA chief executive officer--expect a great deal from the testing system they have in mind.

Besides respecting the nation’s diversity and its constitutionally based tradition of leaving the decisions about education mostly to the states, they want the new tests to enable parents, colleges and prospective employers to get a more thorough, more accurate assessment of an individual student’s knowledge--including thinking ability--than is yielded on most widely used exams.

They want to know how well schools and districts are doing their job of preparing students to live and work in an increasingly complex society. They also want to know how American students measure up to their counterparts in other countries.

Those endorsing a system of national tests also expect them to be a catalyst for schools to provide a more rigorous curriculum and to devise better ways of delivering it to all kinds of students. As Resnick puts it, the national testing system should be “an important lever for really deep changes in the education system.”

“What makes this really interesting is that this is all happening at the same time there is a great revolution about what tests should be and do,” noted UCLA’s Baker, whose research group is trying to evaluate the various types of assessments now in use or being developed.

“One of the ironies of this whole thing is that the inference we are not competitive (with students in other countries) is drawn from tests that have been discredited,” Baker said.

The machine-graded, multiple-choice tests that generations of schoolchildren have answered by coloring in blanks with No. 2 pencils have been increasingly criticized. Although cheap and easy to score, there is growing recognition that the short-answer tests are culturally biased against the poor and minorities, lend themselves to cheating, do not provide meaningful measurements--especially of such higher-order skills as reasoning and creativity--and discourage the teaching of challenging concepts and thinking. Critics also say such tests are harmful because they have often been used to label children and clumsily sort them by purported ability level.

Furthermore, critics say, the results of most traditionally used tests can be easily manipulated to give an overly rosy picture of how well students in a given school are doing. Thus, parents can be assured that their students and their schools are not sharing the achievement problems afflicting many of America’s schools.

More and more states--including California, Connecticut, Kentucky and Maryland--are moving to assessment systems that require students to write essays, conduct science experiments or show how they computed math answers. Vermont is experimenting with evaluating students’ progress through portfolios, or samples of their work collected throughout the school term.

Resnick’s group expects its prototype to produce a multiple assessment that is part written test, part portfolio and part student project.

Although costlier and harder to score and compare, these are the kinds of tests proponents of a national exam system say they want to shoot for.

“We want the American Achievement Tests to be so much a part of everyone’s life that people will be asking each other ‘How did your kid do on the AATs?’ instead of just ‘How did your kid do on the SATs?’ the way they do now,” said Bruno V. Manno, acting assistant secretary in the Education Department’s Office of Educational Research and Improvement. He referred to the Scholastic Aptitude Test, widely used for college admissions.

The Administration has proposed boosting participation in a national exam by offering incentives to states and districts that use the tests and giving presidential citations to students who do well on them.

Development of the exams is just beginning. Alexander has asked Congress for $12.4 million from the coming fiscal year’s budget toward creating standards and tests.

Consensus on what students should be expected to know in every subject and at each of the three grade levels must be reached first in what promises to be a difficult, time-consuming process. At the same time standards are being worked out, a series of determinations must be made about what form the national exam system should take, what legislation might be required, how much it would cost, who would pay for it, and who would oversee it. As yet, there has been no determination as to what groups or agencies will be responsible for developing various aspects of the proposed system, Manno said.

Another issue is what to use in the years before the new system is available. The National Assessment Governing Board, which currently issues an annual report commonly known as “the nation’s report card” on how American students as a group are doing, is working on a plan to develop proficiency standards in several subject areas. Some favor expanding the use of this system’s tests--basically multiple-choice tests given only to a sample of students--in the interim, itself a controversial suggestion.

After months of haggling, the assessment board earlier this month settled on proficiency standards in math for students in grades four, eight and 12. Recently the National Council of Teachers of Mathematics offered curriculum guidelines and standards of its own. Some work has begun on developing science standards, but there has been little so far in English, history and geography.

The national goals panel, headed by Romer, also has been grappling with the standards-setting issue in its efforts to measure progress toward the six national education goals that grew out of the September, 1989, “education summit” convened by Bush with the nation’s governors. The panel’s first progress report is due in September.

That historic summit--the first ever on education--reflected the growing uneasiness about American schools and the apparent lack of progress from a wave of reforms implemented during the 1980s. The governors’ call for common goals--along with various business groups’ interest in getting a handle on what American students can do--gave impetus to the national test movement, considered unlikely until only recently.

Chester E. Finn, Jr., professor of education at Vanderbilt University, a member of the President’s advisory committee and an early advocate of a national testing program, said much of the education Establishment has been slow to recognize a shift in public opinion in the last year or so.

“There are still some dug-in groups, but a lot more people are thinking this is an acceptable idea,” Finn said. “They realize that American education does not have anything resembling an independent auditor, there is not trustworthy advice about whether the books are balanced.”

Finn also said he believes that a single-test system, so long as it remains voluntary, would be politically acceptable should the problems inherent in the multiple-test concept prove insurmountable.

Unless large numbers of schools participate, scores will not be very meaningful for comparison purposes. But educators and policy makers agree there will be enormous pressure on the schools to offer the exams.

A Gallup poll taken shortly after the President unveiled his education reform proposals found 68% of Americans support the idea of a voluntary national test; 65% said they believe such a test would help improve education.

There is still plenty of controversy.

In March, leaders of more than three dozen education and civil rights group called a news conference in Washington to counter the increasing number of proposals for a national exam.

Monty Neill, associate director of the National Center for Fair and Open Testing (FairTest), based in Cambridge, Mass., called such proposals “disastrously misguided” because they divert attention and resources from efforts to improve education.

In an interview after the President outlined his Administration’s proposals, Neill said he supported the kind of multiple-assessment method proposed by Resnick and Tucker, but he said he doubted it could be translated successfully into the national yardstick the Administration has in mind. Once the cost of such is system becomes apparent, Neill said, he fears policy-makers will opt for a single, short-answer test.

“What makes us nervous is not the best intention but that what will result will be something considerably less,” Neill said.

Keith Geiger, head of the National Education Assn., the country’s largest teacher’s organization, said testing is useful only when it helps improve classroom instruction. It is extremely harmful when it is used to compare groups of students and schools, which he said is the usual outcome of “paper and pencil” tests--and a major goal of proponents of a national test. Geiger, like other national test critics, argues the money that the testing system would consume would be better spent in the classroom.

“The whole notion of national tests is a bad idea,” said Theodore Sizer, an education reformer noted for his successful work with children in inner-city schools.

“I am the first one to say that American schools are on the whole mindless and soft,” Sizer said. “I don’t think we lack for good tests. We lack for good schools, and there is no evidence that cranking up the number of tests is a very constructive lever to get better schools.”

Gregory Anrig of the Educational Testing Service and Thomas Shannon of the National School Boards Assn. put aside their earlier opposition based on the general outlines of the Administration’s proposals. It is clear, however, they still have some worries, especially over whether schools will have the resources to make improvements.

“My major concern is not that we can measure something but whether students have the opportunity to be taught whatever it is we decide to measure,” Anrig said. “We’re always telling schools to do a better job at the same time we’re cutting their budgets.”

But Albert Shanker, president of the American Federation of Teachers and a member of the President’s advisory committee, has been an influential voice in the move toward a system of national testing, in part because he feels it can help “restore public confidence in public education.”

“People will not vote to spend money on education if we don’t regain their confidence,” said Shanker. He also said he believes it is possible to develop good methods for developing and administering the tests--but they will be expensive. He pointed to Great Britain’s national exam, which he said costs about $120 per student.

“I’m not saying we should back away from it because it costs money, but should do it right . . . the worst thing we could do is give a test that requires a lot of memorized answers that students will forget the next day.”

Still, Shanker, like many of the others advocating a national testing system, believes it is an idea whose time has come.

“It is the right thing to do,” he said. “It will take time and money, but it will happen. Every other country does this, and ours can, too.”