The Test That Everyone Fears

By John Elson

Seldom has a warning been so baldly ignored. Back in 1926 the nonprofit College Board introduced the Scholastic Aptitude Test with a cautionary observation: "This additional test," said the board, "should be regarded merely as a supplementary record. To place too great an emphasis on test scores is as dangerous as the failure properly to evaluate a score."

So much for caution. In test-happy America, the SAT has since become a kind of academic icon and a national rite of passage for college-bound high school students. Every year more than 1.3 million of them take the 2-hr., 30-min. multiple-choice exam, which is intended to measure students' reasoning skills, math and verbal, as well as their readiness for college. High SAT scores -- perfection is 800 on each half of the exam -- have acquired the cachet of quality. Suburbs lure prospective home buyers by touting the SAT records of their high schools' graduates. Colleges boast of the high average scores of their incoming freshmen; nearly 1,600 U.S. colleges and universities, including the Ivy League elite, rely on SAT results.

Nonetheless, complaints abound. Critics of standardized testing have long charged that the SATs contain built-in cultural and class biases, which put women and minorities at a disadvantage. They claim that the multiple-choice questions do not effectively gauge students' critical thinking skills or their future potential.

Partly to deflect the growing unhappiness, trustees of the College Board last week revised the SAT, although less radically than some critics had wanted. Starting in 1994, the SAT and its companion Achievement Test will be known as SAT-I and SAT-II.The former will test verbal skills and reasoning ability in math; the latter, knowledge of certain specific subjects, such as history and politics. SAT-I will include longer critical reading passages and more questions to test students' understanding of the material.

In the math portion, test takers will have to work out the answers to several problems rather than merely make a multiple-choice selection. They will, however, be allowed to use hand-held calculators. The optional SAT-II will include a 20-min. essay as well as a world-history test.

Harvard University President Derek Bok was chairman of a commission that recommended the changes. The revision, he says, "begins to send a useful signal to schools that problem-solving ability is important, rather than simply the ability to identify the correct answer from a predetermined list." But Bob Schaeffer of the National Center for Fair and Open Testing, a longtime critic of the SAT, charged that the board had failed to deal with the verbal section's analogy problems, which frequently make unconsciously elitist, racist or sexist assumptions about the backgrounds of those taking the test. On one recent test, nearly 16% more men than women were able to select the right analogy to "mercenary: soldier" (hack: writer).

Many educators argue that high SAT scores are no more accurate a predictor of academic success than high school class ranking or grade averages. They also charge that SAT success can be learned, pointing to cram schools that promise, for substantial fees, to raise students' scores by 100 points or more. After a two-year study, Dr. Stuart Katz, a University of Georgia psychologist, concluded last March that the verbal section of the SAT measures test-savvy, not reading ability. He found that 172 college students correctly answered, on average, 38% of the multiple-choice comprehension questions without even reading the test selections. Many colleges, notably in the Midwest, are turning to the rival ACT exam, put out by the American College Testing Program. That 3-hr. battery of exams claims to measure student skills in four curriculum areas: English, reading, science reasoning and math.

The SAT revisions have again focused attention on what one critic, author David Owen, has called "probably the most powerful unregulated monopoly in America": the Educational Testing Service of Lawrence Township, N.J., which prepares the exams for the College Board to administer. And not just the SATs. A nonprofit corporation, ETS is by far the nation's largest private educational assessment service, offering a variety of tests that range from electrology to law to the federally sponsored National Assessment of Educational Progress, which measures student achievement in seven subjects. Founded in 1947, ETS has a serene, campus-like headquarters near Princeton University, a staff of 2,960, more than 270 clients (including the Federal Government), gross revenues of $299.7 million -- and seemingly boundless ambitions.

"Our traditional mission has been to place ourselves at the transitional points of education between high school and college, college and graduate school," says ETS president Gregory Anrig, 58, a former teacher and commissioner of education for Massachusetts. "Now we are expanding into more and more programs that help kids to learn and teachers to teach more effectively."

In fact, ETS is moving as aggressively as a for-profit corporation would in seeking to expand its business. Among other things, the service is using computers and interactive video in new grammar-school courses that are designed to advance critical-thinking skills. In a program currently under development in Brookline, Mass., students play reporters who "investigate" a story from classical mythology under the direction of a computerized editor and their homeroom teacher. Then the youngsters write a story on what they have discovered. ETS is also working with computers to redesign the National Teachers Exam, used in 34 states to license prospective educators. The company's proposed test for fifth-grade math teachers would measure not only the exam takers' knowledge of the subject but their understanding of how fifth-graders learn.

ETS officials deny that the SATs propagate inequity and reflect bias. The fact that upper-income whites score higher than lower-income minorities, they contend, reflects society's imbalances, not the exams'. ETS notes that 400 people of varying backgrounds check every SAT question during 10 review stages and eliminate any that are found to be biased. "To say these tests are biased because results vary," says Gary Saretzky, chief of ETS's sensitivity-review process, "is like blaming the thermometer for the fever."

Perhaps the strongest criticism of ETS is that its zealous promotion of standardized testing emphasizes minutiae at the expense of mind stretching. "In some inner-city schools, kids don't read whole books," says Arthur Weiss, president of the National Council for Accreditation of Teacher Education. "They spend all their time learning how to read three paragraphs and answer multiple-choice questions." Anrig concurs. "You don't need to test a child every three months to see whether he can read," he says. "It's like pulling up a carrot to see if it's growing."

His concern is a major reason why ETS is devoting more of its energies to curriculum development. On tests there may be instant answers, but not in the learning process. It may take years before educators can decide whether Anrig and his colleagues have created new solutions or compounded old problems for America's troubled schools.

CHART: NOT AVAILABLE

CREDIT: TIME Chart

CAPTION: Multiple choices in the old SAT

Open questions in the new SAT

With reporting by Joelle Attinger/New York, with other bureaus