Testy about Testing
Concerns about race, class, unions, and state budget priorities have elevated the testing of students to a major national issue--and muddied the discussion.
Discussions of standardized testing often seem to be about anything but the tests themselves. Advocates of school choice and vouchers hope that poor test results for public school students will convince lawmakers to allow some public money to be used to subsidize private school education. Test results are being used to guide decisions about how funding should be distributed among public schools. Colleges of education are being challenged to release results of how their graduates perform on teacher certification exams. The Supreme Court has agreed to review the question of affirmative action because of complaints that some minority students with low standardized test scores are being admitted to selective state universities. With so many powerful and emotionally charged issues swirling around, it is often difficult to focus on the tests themselves.
One of the ironies of standardized testing is that it began as a way of opening the doors of the elite colleges to the underprivileged. Harvard University president James B. Conant believed that admissions officials gave the graduates of expensive private schools an advantage over public school graduates in gaining admission. He wanted a reliable and objective way of comparing the achievement of all students, believing that a vast number of public school students would be able to demonstrate their academic superiority. Some of his benighted counterparts welcomed the idea of testing for the opposite reason; they believed that the test results would give them a justifiable rationale for excluding certain religious and ethnic groups. Conant was right, and a diverse group of high-scoring students began to gain access to the best schools.
Trouble began when it became clear that standardized testing was not the solution for all groups. In particular, African-Americans, whose average tests scores were lower than those of other groups, began to see standardized testing as a barrier to admission. Universities that were eager to have a fair representation of African-American students either set different requirements for standardized test scores or gave more weight to other factors, such as overcoming adversity, in making admissions decisions. Applicants who consider these admission practices unfair have taken the universities to court, and now the practice will be considered by the Supreme Court.
One of the stated purposes of President Bush's No Child Left Behind legislation is to ensure that poor and minority students are being educated to meet the same standards as all other students. The goal is to make all schools accountable for providing an adequate education to all students, but critics fear that poorly funded schools with large numbers of low-income and minority students will be punished by reductions in funding if their students score poorly. They say that the legislation has unrealistically high expectations for improving performance of poor and minority students in the short term and that it is unfair to sanction schools who are not meeting expectations quickly. They fear that the legislation could be implemented in a way that punishes students for the misfortune of attending underperforming schools rather than using the tests as a tool for improving education. The legislation aims to enable students at underperforming schools to transfer to more successful schools, but for many students and school systems this is not a practical option, because students are reluctant to leave their friends, the good schools are already overcrowded, or transportation is inconvenient.
Teachers themselves have doubts about testing. They worry that the focus on a single test will narrow the curriculum and lead teachers to emphasize test-taking skills at the expense of more important matters such as writing and problem solving. Why can't we simply trust teacher evaluation of students? One reason is that too many teachers have abandoned their responsibility to assess students rigorously. The College Board reports that almost 40 percent of students who take the SAT exam have "A" averages, but colleges find that a growing number of their students need remedial classes because they have not learned the basics needed for college. Having encountered college freshmen who do not know the difference between upper and lower-case letters, I have no doubt that at least some schools and teachers need to be held more accountable.
Even real estate agents have a stake in the test results. Higher scores mean higher home prices, and housing and schools have a symbiotic relationship. The children of the affluent score higher on average than those of lower income groups, and higher home prices mean more affluent homeowners and higher scores. In addition, many schools are funded out of real estate taxes. Higher home values mean more revenue for the schools and the opportunity to improve the quality of education. Of course, the situation is the reverse for students from low-income families, and the achievement gap between rich and poor seems destined to grow.
Back to basics
With all respect for the wisdom of judges, teachers' union officials, school choice activists, real estate agents, and all the others who have strong opinions about testing, we thought it might be enlightening to listen to what educators, psychometricians, and learning specialists have to say. In the interest of clarifying the discussion of testing, the University of Texas at Dallas and the National Academies sponsored a symposium on student assessment in February 2002 with the goal of returning the focus to testing itself. The symposium paid special attention to the experience of Texas, which has been a leader in testing and a model for national policy. The articles in this issue began as presentations at that conference. They examine the thinking behind national policy on testing, the science that should guide testing, the problems that can emerge when testing is misused, and the testing industry's capacity to maintain quality as the demand for testing grows.
We do not have room for articles by all the speakers at the symposium, but for those who are interested we want to recommend the excellent work done by the other speakers: Texas education reform activist Darvin Winick, Dallas school superintendent Mike Moses, National Research Council staffers Michael Feuer and Pat De Vito, Amherst College economist Steven Rivkin, Just for the Kids executive director Brad Duggan, University of California at Berkeley psychometrician Mark Wilson, and University of Texas at Dallas education policy analyst John F. Kain.
Testing is important, and there is much about testing that needs to be debated. But the debate should begin with the details of testing itself rather than peripheral issues. As a first step, we should dismiss the categories of "pro-testing" and "anti-testing." The vast majority of Americans see testing as a useful way to ensure that students are learning what we want them to learn. The discussion of testing needs to be about whether the full range of desirable skills is being assessed, whether the standards of performance are fair and reasonable, whether the test is aligned with the curriculum, whether test results are being considered in context with other indicators of student achievement, whether test results are being properly interpreted and used for the purposes for which they were designed, and whether testing is being used for its ultimate purpose: to gain a better understanding of how well young people are learning and to gain insights into what can be done to enhance their education.