New Book from The Civil Rights Project Highlights the Limits of Test-Driven Reforms
A new book from The Civil Rights Project at Harvard University finds that the current overreliance on high-stakes testing threatens to deepen America's educational inequities. The book, Raising Standards or Raising Barriers?, is co-edited by Gary Orfield, co-director of The Civil Rights Project, and Mindy L. Kornhaber, the project's research director for elementary and secondary education. It is published by the Century Foundation Press as part of a series called Civil Rights in a New Era. The book makes clear the importance of high standards and accountability systems. But support for standards and accountability systems should not be equated with support for high-stakes tests. These are tests that are used to determine whether a student graduates, gains access to challenging curriculum, or is promoted, or whether schools or educators are rewarded or penalized. Most of the contributors to the volume have found evidence that policies that focus on high-stakes testing corrupt educational reform and undermine achievement, especially for at-risk students. State and federal policymakers are increasingly pushing such tests as a panacea for the nation's educational concerns.
Orfield noted, "Congress has just voted to greatly expand mandated state testing, requiring the development of 213 additional state tests, the expenditure of billions of dollars, and the loss of a great deal of instructional time in schools across the country. The time has come to hold the testing industry accountable and to ask what do we really know about the costs and benefits of testing. Too often we are getting poor tests which are misused to test things not taught and to punish the children attending inferior schools, violating basic concepts of fairness and civil rights."
* Policymakers assert that testing will improve the national economy. Yet, economist Henry Levin of Columbia's Teachers College writes that test results "show only nominal statistical links to measures of worker productivity." Given such weak links, an economic system that depended on test results for job selection would be massively inefficient: it would simultaneously reject many capable applicants and accept many poor performers.
* Policymakers argue that high-stakes tests will improve motivation. Yet, George Madaus and Marguerite Clarke, of Boston College, assert that policymakers have paid little attention to "who will be motivated and who will not?" Students at the margins of performance tend to "dismiss the examination because they feel they lack the ability to do what is necessary to pass." Despite much rhetoric, it is not evident that testing motivates students who are already at risk.
* Policymakers claim that high-stakes testing will improve teaching and learning. Yet, in their research on schools serving predominantly minority and poor students in Texas, Linda McNeil of Rice University and Angela Valenzuela of the University of Texas at Austin find that score-raising tactics often replace sound teaching. For example, curriculum is reduced to test preparation, funds are directed toward commercial test prep materials instead of library books and laboratory equipment, and instruction in untested subjects - including science and history - is minimized. In such schools, researchers say, "even the most knowledgeable teachers are asked to set aside their lesson plans and materials to teach to the TAAS." While Texas has shown gains on the National Assessment of Educational Progress (NAEP), it makes little sense to generalize from Texas' policies to the nation as a whole. According to a nation-wide analysis by Monty Neill, executive director of the National Center for Fair & Open Testing (FairTest), states with high-stakes tests do not show a clear pattern of improved performance on the NAEP relative to states without such tests.
Alongside noting effects on teaching and learning, the contributors to this volume illuminate other troubling consequences of high-stakes testing policies.
John Bishop of Cornell University, a contributor and supporter of high-stakes exams, nonetheless notes that dropout rates are likely to rise under New York's new Regents exam policy. Gary Natriello and Aaron Pallas, of Columbia's Teachers College, show that under high-stakes testing policies in New York, Texas, and Minnesota poor and minority students will be less likely to receive a high school diploma.
# Retention rates: The results of high-stakes tests are increasingly being aimed at "ending social promotion." University of Wisconsin Professor Robert Hauser finds that, "Test-based promotion policies are likely to raise … costs of schooling without corresponding educational benefit." Hauser's analyses show that retention rates are already high, especially for male students and African Americans. At the same time, retention is ineffective: students who are retained ultimately learn less than students with similar test scores who are promoted. Being left back also greatly increases the risk of dropping out of school.
# Test use should be guided by standards set forth by testing professionals. Jay Heubert of Columbia's Teachers College asserts the need to incorporate existing professional standards for testing into policies that call for students to be tested. For example, the Joint Standards produced by the American Educational Research Association, the American Psychological Association, and the National Council for Measurement in Education, say that a single test should never be the sole basis for making a high-stakes decision. Linking professional testing standards into existing laws and regulations governing civil rights and education would aid the enforcement of appropriate test use.
# Test results should be put to constructive uses. Robert Hauser states that test results should provide feedback that enables educators to discern patterns of strengths and weaknesses in students' knowledge and skill. Results for individual students should be used for early diagnosis and intensive early intervention, rather than simple retention. For this to happen, test data should yield clear and timely information. Test data should also be linked to resources aimed at strengthening classroom instruction and learning.
# Test results should reflect back on the entire system of education. Gary Natriello and Aaron Pallas suggest that test scores should provide policymakers with data for understanding educational processes and opportunities and for making educational resources more equitable. Any stakes associated with testing policies should be shouldered not only by students and educators, but also by policymakers who have control over the educational systems in which teaching and learning occur.
Raising Standards or Raising Barriers? has been published at a time when the federal government is poised to embark on a massive expansion of student testing. The U.S. House and Senate have just passed education bills that incorporate George W. Bush's proposal of annual testing for all students in grades 3 through 8. Because test scores will be used to penalize low-scoring schools, they will act as high-stakes tests for teachers and administrators especially in schools serving high proportions of poor and minority students. "Unfortunately, there is simply no evidence that efforts to raise test scores will provide poor, minority, and bilingual students with the kind of high quality education that their more affluent counterparts receive," said Mindy L. Kornhaber, the volume's co-editor. "Instead of devising testing schemes, which commonly lead to more retention and dropping out, policymakers should focus on providing all students with appropriately certified teachers, challenging curriculum, high quality facilities, and ample supplies of up-to-date books and materials."
About The Civil Rights Project
The Civil Rights Project at Harvard University (CRP) is an interdisciplinary think tank whose central mission is to mobilize the resources of Harvard and the broader research community in support of the struggle for racial and ethnic justice. By building strong collaborations among researchers, community organizations, and policy makers, CRP hopes to reframe the tone and content of many of the current legal and political debates, and to support the work of others around the country. Since its founding in 1996, CRP has commissioned over 75 studies related to civil rights and educational reform.
For More Information
Contact Christy Hicks at the Century Foundation, 212-452-7723