National Tests: A Yardstick to Learn By
By William J. Bennett and Chester E. Finn Jr.
One view has been voiced by some conservatives and Republicans, notably Rep. Bill Goodling, chairman of the House Committee on Education and the Workforce. They oppose national testing under any condition.
The other view is that of the Clinton administration, which has its own national testing scheme.
We consider both options insufficient and have a third alternative that would make fundamental changes in the Clinton plan to rescue a good idea. These changes, somewhat modified in negotiations with the White House, were introduced in the Senate by Sen. Dan Coats and passed 88 to 12 last week. The House, meanwhile, is expected to turn thumbs down on national testing.
Conservative critics argue that a new national test would provide no new or useful information, would lead to a national curriculum, would reduce local and parental authority and would fail to generate higher academic performance.
We too have deep concerns specific to the Clinton proposal. However, we believe many arguments made by conservative anti-testers are substantively misguided and politically unwise. To those who make a philosophical argument against the very idea of national tests, we would point out that almost every serious education expert Democrat or Republican, liberal or conservative will testify to the value of today's federally funded National Assessment of Educational Progress (NAEP) tests. They hold NAEP in high regard because it provides important information, its content is sound and it has rigorous standards built in in large part because the NAEP is governed by an independent board.
By law, however, the NAEP cannot be used by districts or schools to report the progress of individual students. In other words, it can tell you how one state is doing relative to another state, but not how well one district or school is doing relative to another. We advocate extending voluntary NAEP-based tests to school districts, schools and students. Absent that, we will continue with a system that does not allow for standards-based tests of core academic skills that can be used at any level, by any "consumer" (governor, parent, local school board or charter school principal) to yield clear information about comparable achievement.
For example, today's "standardized" tests do not actually contain any standards. They merely provide a statistical average: where you are within the group. Testing "at or above grade level" merely indicates how a student is performing relative to other students, not how much the student knows. And performing "at or above grade level" is not encouraging if the grade-level bar is set low which is now the case.
As for the argument that voluntary national tests would reduce local and parental authority: If they are done right, the opposite will occur. The tests will empower parents by providing them with information critical to the success of reforms such as charter schools and school choice. And keep in mind: these tests would be voluntary. We are a nation at risk and our flawed testing regime is one reason why. A good voluntary national test would be an important step forward. A bad testing program, on the other hand, would be a giant step backward, which brings us to the Clinton proposal.
In a brazen power grab, the administration has attempted to promote its version of a voluntary national test without seeking authorization from Congress. The Clinton plan would hand the test-development process to the very kinds of organizations and groups that have helped ruin modern American education. It would dumb down the test standards and reflect the "trendiest" read: "worst" ideas in education, things such as "whole language" reading and "constructivist" math. The result would be tests that wouldn't indicate whether students could read or do math. The Clinton plan would remove responsibility for national testing from the National Assessment Governing Board (NAGB) a well-respected, independent, nonpartisan body created to set policy for NAEP and give de facto control to the Department of Education.
The Clinton plan would politicize and subvert the whole idea of national testing and has flaws so misguided as to be dangerous. This explains why only six states have signed up for it and the governor of one of those states, Michigan's John Engler, says he will jump off this train unless fundamental changes are made.
If faced with a choice between no test (the Goodling position) and the Clinton test, we would endorse no tests. Fortunately, however, the common-sense Coats proposal makes key changes to the Clinton plan, which include:
Giving the National Assessment Governing Board control over all policy aspects of the national tests.
Giving the NAGB authority to review and change the national test specifications, development contracts and advisory committees already launched by the Education Department. Otherwise, bad decisions and contracts already made would stay in place.
Charging the NAGB with ensuring that the content and standards of national tests be the same as those of the National Assessment of Educational Progress. Otherwise the public would not know whether the test results meant the same thing as NAEP results.
Making key changes in the composition of the NAGB so that it has greater bipartisanship and independence. At present, just two of its 25 members need to be from the party out of power, and given the huge responsibility of national testing and the need for bipartisan consensus, this imbalance is wrong.
We prefer no test to a bad test. But most of all, we prefer a good test one with integrity, high standards, solid content and procedural safeguards to insulate it from politics and partisanship. That is what the Coats amendment would provide and why we hope that is the legislation that will emerge from the Senate-House conference committee. It is legislation that reform-minded legislators of both political parties should embrace and one which the president should sign.
The writers are, respectively, a former secretary of education in the Reagan administration, and a John M. Olin fellow at the Hudson Institution, a nonprofit research organization.
© Copyright 1997 The Washington Post Company