Nationwide Student Testing? Here Are Some Answers

The upside of proposed national reading and math testing

Should fourth graders across the country be tested annually on their reading ability and eighth graders on their mathematics knowledge? It seemed a no-brainer when President Clinton proposed it last February. Yet in recent weeks the notion of national testing has become highly politicized. The initiative, which passed in the Senate, has run into problems in the House, and the debate over whether and how such testing should be funded is likely to rage on. Some opponents contend that national testing--even though voluntary--would be the first step toward a national curriculum and federal control of the nation's 15,000 school districts. Others argue that poor and minority students would be unfairly penalized for low scores. Below, some key questions and answers about national testing:

What would national testing of fourth and eighth graders accomplish?

The tests would provide parents, teachers, and students with individual results that can be compared with district, state, national, and even overseas scores. Business in particular favors testing as a way to benchmark the academic achievement of U.S. students--against one another and against students in other countries. Today there is no way to compare the performance of a child in Utah with that of a child in Kentucky, much less with that of a child in Tokyo or Taipei.

In the past few years, many states introduced or upgraded academic standards. National tests, then, would "help us know whether what we're doing is good enough," observes Governor Roy Romer (D-Colo.). "National tests would provide a benchmark against which we can measure progress and be a diagnostic tool to tell us where to concentrate our efforts."

How would the tests be designed and conducted?

National tests would be based on the only existing standardized national test, the National Assessment of Educational Progress. Why not just use NAEP? Because it doesn't provide results for individual districts and students, it's not conducted annually for each subject, and because its sample size of about 100,000 students means it's not representative.

In August, the Education Dept. received draft reports and sample questions from national test panels (table). The department has already contracted with several test developers to design the tests and estimates that design and field testing will cost $16 million. Actual testing could total $100 million in 1999, or about $10 to $12 per child. The National Assessment Governing Board, an independent body that currently oversees NAEP, will have oversight.

To date, only seven governors or state school superintendents and 15 municipalities have said they will participate. A number of governors have pushed in recent years for tougher standards and assessments within their states, and they may regard the President's initiative as an unwelcome distraction.

Don't kids take enough tests already, and won't national tests promote "teaching to the test"?

There are some states where students are subjected to a battery of state tests and others where students take very few. Today, there certainly aren't too many national tests being given, and if the new tests duplicate state-administered tests, perhaps some of the latter could be eliminated. As to "teaching to the test," if national tests are well designed and a good tool for assessing students' skills, there shouldn't be a problem. "If a test is measuring what we want students to know, it's not a bad thing to teach to the test," observes James W. Stigler, an educational psychologist at University of California at Los Angeles.

Will national testing lead to better learning?

Possibly, but only indirectly. "Testing in and of itself doesn't do much of anything to improve learning," says Linda Darling-Hammond, a professor at Teachers College at Columbia University. South Carolina and Georgia ramped up statewide testing in the 1980s but have experienced little or no improvement in educational achievement, while states such as Iowa and Minnesota do relatively little testing and rank very high in educational achievement. One factor: Iowa and Minnesota have tough standards for teacher education and licensing. "Tests don't teach children, teachers do," says Darling-Hammond.

So if national testing leads to better teacher training and stronger curricula, better learning will result. That means a big investment in teacher training. Today, 30% of math teachers don't even have minors in math. More than 50% of students taking physical science are taught by a teacher with no background in that field. "We've got to give kids qualified teachers--we can't put the whole onus on kids and tell them to learn to pass tests themselves," says Darling-Hammond.

Will national testing lead to a national curriculum?

Opponents such as Representative William F. Goodling (R.-Pa.), a former educator, worry that national tests could be a first step in that direction, and they believe curriculum decisions should be made at the local and state level. Indeed, the politics of education are such that control could never be wrested away by the federal government, nor is there any indication that the federal government would want to do so. Yet it's true that national tests would presume that there is some single standard of reading ability and mathematics knowledge against which students are being measured--an unstated national curriculum, if you will. It hardly seems objectionable, however, to expect fourth graders to read at the same level from state to state or eighth graders to have the same mastery of math skills from coast to coast.

If states are already moving to improve education, do we need national testing?

States are doing better, but it could take many years before there is convergence toward higher academic standards. Currently, state NAEP results show much lower academic achievement than does state-developed testing. National tests would put more pressure on states to improve. But Clinton's proposal shouldn't be viewed as a magic bullet for the nation's educational ills. Participation is voluntary, so truly nationwide results wouldn't be available. At best, national testing could turn a spotlight on students' performance and perhaps motivate parents, educators, and employers to push for better results. How to go about getting those results is part of a bigger and still more contentious debate.

Before it's here, it's on the Bloomberg Terminal.