Ever since the publication of A Nation At Risk in 1983, the standards-based school reform movement has attracted a broad coalition of educators and politicians from both the left and the right. Conservatives applaud its emphasis on accountability, while liberals and civil rights advocates are drawn to its underlying premise that all children can learn at high levels.

Unfortunately, this movement has been diluted in many states to one policy: high-stakes testing. Under this policy, the score on standardized tests is linked to grade promotion, high school graduation, and, in some cases, teacher and principal salaries and tenure decisions. George W. Bush, has called such testing the "cornerstone" of all educational reform.

Despite its current political popularity, a growing body of research suggests that the testing "solution" may be more onerous than the problems it originally sought to alleviate. Major findings reported during the past year include:

High stakes tests attached to high school graduation lead to increased drop-out rates, particularly for poor and minority students. This phenomenon has been documented by researchers Walter Haney and Gary Natriello in Texas, where high stakes testing policies have been in place for some time. Designs for Change and the Gaston Institute at the University of Massachusetts have released evidence that new testing policies are almost certain to significantly increase dropout rates in Chicago and Massachusetts. Minority students are heavily overrepresented among students who fail to receive diplomas because of standardized testing requirements.

Using tests to retain students in the same grade increases the likelihood that the retained student will eventually drop out. This was the conclusion drawn by Robert Hauser, who also found that the practice of retaining students in grade produces no educational benefits, increases classroom management problems, and is tremendously expensive for school systems.

High stakes tests dilute the curriculum. Two thirds of over 1,000 teachers surveyed recently reported they felt far too much time was spent teaching to the test "to the detriment of important areas of learning." Moreover, a "teaching to the test" approach is far more likely to dominate instruction in high poverty schools than in affluent ones. For instance, researchers in Texas found that test preparation has replaced much of the curriculum in many poor schools in Houston.

Many states are using tests in ways that directly contradict the recommendations of experts and the federal government. National testing experts, the National Research Council, the American Educators Research Association, the American Psychological Association, the National Council on Measurement in Education, and the Department of Education all assert that no decision of serious consequence in a child's life should be made on the basis of a single test score. Yet many school districts are doing exactly that, as well as using tests designed for other purposes to make tracking, promotion, and graduation decisions.

If we are serious about raising standards, we should implement reforms that dozens of research studies have shown to be effective: expanding access to quality early childhood education, reducing class size, increasing the pool of skilled and committed teachers, upgrading instructional facilities and designing assessments that are linked to appropriate and timely interventions. Within such a broader scope of reforms, standardized tests can be useful tools for identifying curricular weaknesses and targeting students in need of additional support. Without these, high stakes testing policies tied to harsh sanctions for students are likely to further exacerbate, rather than close, the achievement gap between poor and affluent students in this country.