High Stakes for Education - Swarthmore College Bulletin

As President George W. Bush’s education reform proposals limped through Congress this year, their reliance on standardized tests to determine the fate of students, teachers, and schools has sparked a national education debate. From the classroom to the school district office to the Department of Education in Washington, D.C., federally mandated testing raises the stakes for kids--and a lot of questions for educators.

Although Bush lost the legislative battle to provide taxpayer-financed private school tuition for children in low-performing schools, he won the fight for expanded testing. Starting in the fall of 2004 (House bill) or 2005 (Senate bill), all public school students in grades three through eight will be tested annually. High school students will also be tested at least once in reading and math, and the Senate bill also requires testing in science. “Failing schools” will face penalties unless they improve.

Following national guidelines, states that don’t have their own tests will be required to adopt them, continuing a trend that has been accelerating at the state and local level for years. And although nothing in the federal law says the tests should be used to evaluate individual students, that also may happen.

The Bush plan has garnered bipartisan political support, but many teachers, parents, and educational theorists have loudly protested the theory--for it remains a theory, with spotty success and much failure--that frequent mandatory testing will lead to improvements in public education. Federally mandated testing implies that results can be measured against national standards that are difficult to evaluate--or even agree upon--in America’s highly decentralized educational system.

Increasingly, states have begun their own testing programs. Even in states without testing, cities such as Philadelphia, New York, and Chicago have adopted their own systems in an effort to stop automatic social promotion. In Philadelphia, students are told that their scores on citywide proficiency exams in English, math, and science will constitute one-sixth of their final grades. Alexandra Volin ’96, who has taught high school English in Philadelphia’s Parkway Program for four years, believes that such testing “ends up taking huge significance. It depresses the students, who stop believing that anything we do [in the classroom] really matters.”

In any case, says Volin, “in our school and several others, we didn’t get the scores in time to factor them in,” even though the purpose is to align grades across the district. High-stakes testing, she believes, “is about making districts and schools fall into line. It’s a political bludgeon.”

For the classroom teacher, Volin says “the issue is about how much time is taken out of classroom time for these tests.” She ticks off five instances over the course of the year when the school schedule was interrupted for testing, with makeup tests given three to five days later. Volin has learned that “it’s not enough to close your classroom door and teach or to have a sense of how education ‘should be’ in America. You have to be committed to school- and school district-level activism, building relationships with peers and advocating for change.”

One of her current projects is critiquing high-stakes testing in the Philadelphia district. With another teacher, she submitted a letter to the Board of Education and attended its meeting in May. The board requested a testing liaison to improve testing conditions at the school. A meeting with the district’s chief academic officer was scheduled for midsummer.

Contacting the “Real World”

Volin’s concern is shared by Professor Eva Travers, director of Swarthmore’s Program in Education, who has taught Urban Education at the College for more than 20 years. Many aspiring teachers have taken the course, which includes regular visits to schools in Philadelphia and Chester, Pa. Last spring, some 25 students participated. “In both districts, they saw the effects on students and teachers of spending two months preparing for standardized tests,” Travers says. “For the last two or three years, many of the students’ journal entries have reported overwhelming preparation for standardized tests, especially at the elementary school level.... The results for teaching and learning are, to say the least, disruptive.”

One Urban Education student, Dan Consiglio ’03, visited Alex Volin’s 10th-grade classes in Philadelphia. “It was one of my few contacts with the real world,” he laughs. Going on school visits “took me out of my college mode of thinking.” At Parkway, he enjoyed noting the different ways in which the two almost entirely African-American classes responded to the same lesson plan as they read Alice Walker’s novel The Color Purple. Conversations with Volin at day’s end centered on issues such as making curriculum choices and dealing with disciplinary problems.

Back at Swarthmore, regular class discussions give Travers’ students a good idea of what their classmates have experienced and provide examples of “what we read about for class,” Consiglio says.

One of Consiglio’s most vivid Parkway observations--which he recorded in his journal and summarized in a final paper--concerned a standardized math test for which calculators were to be provided. At test time, many of the needed calculators remained locked in a cabinet for which no one had brought a key. The students had to start late and complete the exam the next day. “As ‘standardized’ as these tests are supposed to be,” he observed, “it depends on which school you’re in. When I took standardized math tests in my own school, all the students were provided with calculators.”

The no-calculator incident, with its “completely nonstandard” time frame, strikes Alex Volin as a prime example of “irregularities in the way tests are administered.” The students were also “understandably confused” when presented with tests labeled “Algebra I” when they had taken “interactive math,” which combines problem solving in algebra, geometry, and trigonometry. Their math teacher had not been given enough information to gear her teaching to the test even if she had wanted to. “They failed in droves,” Volin says.

Will Testing Improve Learning?

The term high-stakes testing has no absolute definition, but here, it refers to standardized tests whose results are used to determine a student’s promotion from one grade to the next and/or graduation from school. Test results that are used to reward or punish schools or teachers reflect a similar concept: accountability.

It has long been assumed that teachers like Alex Volin know best which students are ready to be promoted or graduated. Only the teacher sees a continuum of student performance and motivation throughout the school year. Yet because teacher evaluations can be subjective and variable, some believe that standardized tests alone should decide which pupil should attend summer school, be held back a grade, or be refused a diploma. Critics of testing argue that such standards apply a near-robotic concept to student evaluation and represent a loss of faith in the individual teacher--long the foundation of education.

In attempts to improve schools swiftly (and, some say, cheaply), politicians ranging from President Bush to local boards of education have seized upon standardized tests that allow comparisons among students, teachers, schools, and school districts.

“Increasingly, political leaders make the granting of more funding for education contingent on evidence that students are actually learning more, often as measured by standardized tests,” explains Jay Heubert ’73, associate professor of education at Teachers College-Columbia University and adjunct professor of law at Columbia Law School. “Some people see tests as a quick fix for low achievement,” Heubert points out, “in part because scores typically go up in the first few years after new tests are introduced. But in themselves, tests don’t improve learning any more than a thermometer reduces fever; they’re just gauges.

“The key is to use information from tests and other sources in ways that lead to improved teaching and learning,” Heubert continues. “Such improvements require training, time, and resources--and they rarely happen within an election cycle.”

Most educators’ preferred remedies for poor student performance include long-term efforts such as instituting smaller classes, paying teachers more, improving physical facilities, supplying better instructional materials, and attracting experienced teachers to low-performing schools. Standardized testing is only one component of the education reform movement, and ironically, it is overtaking K-12 education at the same time that some higher-education decision makers are de-emphasizing standardized tests such as the SAT as admissions tools.

Given the debate over high-stakes testing and its potential consequences for students, Congress asked the National Academy of Sciences to conduct a study on appropriate, nondiscriminatory use of tests for student tracking, promotion, and graduation. Heubert served as study director and co- editor of the resulting report. Published in 1998 as High Stakes: Testing for Tracking, Promotion, and Graduation, the report has influenced national and state debates over promotion and graduation testing. For example, the U.S. Department of Education’s Office for Civil Rights drew heavily on the report in its December 2000 resource guide on high-stakes testing, as did U.S. Senator Paul Wellstone (D-Minn.) in his proposed federal legislation on appropriate use of high-stakes tests.

An End to Social Promotion?

Since 1999, the number of states requiring graduation tests has increased slightly, from about 18 to 20, but promotion testing is another matter. In response to concerns about “social promotion”--promoting low-achieving students so they can remain with their age peers--states and school districts are increasingly requiring students to pass tests as a condition of being promoted to the next grade.

The argument for ending social promotion, which has been expressed by both presidents Bill Clinton and George W. Bush, is based, in part, on the assumption that most schools retain very few students. Research for High Stakes, however, found that as of 1996, even before the latest debates over social promotion had started, retention rates were fairly high nationwide, with big differences by gender and race.

“By ages 15 to 17,” Heubert says, “about 20 percent of all girls are at least a year behind most students of their age compared with about 50 percent of black and Latino boys.”

Despite these statistics, the growth of promotion testing is a real problem, according to Heubert. After reviewing decades of research, the High Stakes authors found that low-performing students who are promoted to the next grade “actually do better academically, socially, and in terms of dropout rates than similar students who must repeat a grade. The strongest predictor of who drops out is who is retained or ‘held back,’” Heubert says. “Dozens of studies and meta-analyses over 20 years show that retention in grade is bad for the kids who are retained.”

Heubert notes, however, that the serious consequences of retention are often not apparent until years later because many children are too young to drop out when they are first held back. “In this sense,” he says, “retention is a silent threat, like high blood pressure.” As a result, the rapid growth of promotion testing, especially in large cities, is likely to create an increasingly large class of students--disproportionately consisting of minority students, immigrant children, students with disabilities, and disadvantaged students--who are at increased risk of dropping out because they were held back once or more.

Proponents of standards-based reform and high-stakes testing point out that these student groups are most often educated poorly and would, therefore, have the most to gain if all schools, teachers, and students were held to high standards of teaching and learning. Critics fear that large numbers of such children will disproportionately be retained in grade or denied high school diplomas because their schools have failed to expose them to the knowledge and skills needed to pass the tests.

A second National Academy of Sciences report, slated for publication in the fall by the Committee on Educational Excellence and Testing Equity, “will stress that dropping out of school is a process that starts early,” says eminent psychologist Ulric Neisser M’53/H’98, professor of psychology at Cornell University and one of the scholars who signed off on the report. “Overall,” he adds, “we don’t yet know whether high-stakes testing will increase the dropout rate. We don’t have good data yet, and the tests are too new.”

The best solution, according to High Stakes, is early intervention: identifying and addressing problems of low achievement before students take promotion tests.

Another idea behind testing on a national level is that results from California could be compared with those from Texas or New Hampshire. Now, says Eva Travers, “every state does it differently.” Yet because states that don’t already test will have to design or purchase their own instruments--and because the federal government has no mechanism for creating tests--there is no guarantee that results will be consistent or comparable.

“High-stakes testing has a bad name because many districts buy off-the-shelf tests that have no relation to what they are teaching,” says Neisser. “But politicians like these tests because they are inexpensive. It’s expensive to hire more teachers and improve school buildings.”

There are other problems, too, such as fraud. Because teachers are held accountable for their students’ test results, Travers says, some cheat by giving students practice in the test questions ahead of time. To keep scores high, some teachers ask low-performing kids to stay home on the day of a test.

Teaching to the Test

The perceived need to spend a great deal of class time on test preparation frustrates many educators. Yet, as Heubert points out, “‘Teaching to the test’ is not an undesirable, unforeseen side effect” of high-stakes testing. States adopt achievement standards and tests precisely so that teachers will focus more on the knowledge and skills that the tests measure. Whether this is a problem depends partly on how good the tests are and partly on how well the students were being taught before.

Not all standardized tests are unsatisfactory, agrees Eva Travers. Some, such as those requiring writing samples, “provide more authentic assessments than do those with multiple-choice answers,” she says.

The way in which the results are used makes a critical difference. Hard work and plenty of communication and cooperation are extremely important. Among the strongest recommendations in High Stakes is that “accountability for educational outcomes should be a shared responsibility of states, school districts, public officials, educators, parents, and students. High standards cannot be established and maintained merely by imposing them on students.”

But quality costs money. “Additional financing and resources are prerequisites for using high-stakes testing to evaluate students for promotion from grade to grade or for graduation, to evaluate teachers, or to evaluate whole schools,” Travers says. “High-stakes testing presumes that you can bring the bottom up, but it’s ludicrous if you don’t provide resources and support.”

Many believe class size makes a profound difference in performance. “Experiments show that the student-teacher ratio matters,” Neisser says, “especially for blacks.”

In December 2000, Congress appropriated $1.6 billion for the 2001-2002 school year for the federal Class-Size Reduction Program. All states are receiving federal funds to recruit, hire, and train new teachers, especially in grades K-3. Since the early 1980s, at least 20 states have begun their own class-size reduction efforts, according to an article in Education Week (June 27). But professional development and other strategies might be more beneficial than hiring legions of underqualified teachers solely to decrease class size, the article notes.

“It would also be helpful to have ways of ensuring that high-stakes tests are used properly,” Heubert points out. Unfortunately, as concluded, the current mechanisms--professional discipline and legal enforcement--remain inadequate. The testing profession has no procedures for monitoring test use or reviewing complaints; courts have few specific standards to enforce.

The Future of Testing

Travers believes that high-stakes testing “will continue because of public and political pressure for accountability until it’s obvious that the process is failing.” For a cautionary tale, one might look to Virginia, where 80 percent of students failed one part of graduation tests implemented several years ago. Heubert reports that if the most demanding state graduation tests were used to deny diplomas today, failure rates would be about 40 percent for all children and 75 to 80 percent among minority, immigrant, and disabled students.

Some states that have favored high-stakes testing are modifying their requirements. In June, North Carolina chose to reduce testing time in favor of instructional time. Arizona has decided to temporarily delay the date by which students must pass the state exam to receive a diploma. Wisconsin has decided to use grades and test scores (rather than test scores alone) to determine which students will receive diplomas. New York has created a waiver, allowing students with disabilities to receive local high school diplomas even if they don’t pass the more rigorous state Regents’ exams.

Neisser thinks that “we have to educate politicians and the electorate about the difference between doing things right and doing them fast. It will take some time, and it will not be easy, but it’s not hopeless.”

Heubert concludes: “The bottom line is that these tests should be used to leverage improved learning and to give teachers and students the help they need so they can succeed. The danger is that the tests may be used to punish kids for not knowing what we’ve never taught them.”

Marcia Ringel is a writer and editor in Ridgewood, N.J.