U.S. Still Lags in Math, Science, Tests Find : Education: Exam officials say results should give Americans a starting place to set standards. But criticism is growing that such comparisons aren’t useful.

Amid growing debate over the usefulness of international comparisons, the Educational Testing Service on Wednesday released a 20-country survey that shows American schoolchildren continuing to fare poorly in math and science.

In the second International Assessment of Educational Progress, American 13-year-olds scored lower in both subjects than their counterparts in most of the other countries surveyed, while 9-year-olds ranked third in science but placed near the bottom in math.

On tests conducted last year, in which each country’s students were rated on the percentage of questions they answered correctly, American 9-year-olds gave the right answers on 58% of the math questions and 65% of the science questions. The top scores of 75% in math and 68% in science were earned by South Korean students. Average scores earned by the 3,300 students in the participating nations were 63% in math and 62% in science.

American 13-year-olds scored 55% in math and 67% in science, compared with the top marks of 80% by the Chinese in math and 78% by the South Koreans in science. Average scores were 58% and 67%.

In science, American 13-year-olds were outscored by test takers in Canada, England, France, Hungary, Israel, Italy, South Korea, Scotland, Slovenia, the countries that made up the former Soviet Union, Spain, Switzerland and Taiwan. They tied with students in China and did better than those in Brazil, Ireland, Jordan and Portugal.

Japan, Germany and some other keenly competitive nations decided not to participate in the survey. Nonetheless, the results added considerably to mounting evidence that it will be very tough to meet President Bush’s goal of making American students first in those subjects by the turn of the century.

Gregory R. Anrig, president of Educational Testing Service, said the survey, should give Americans a starting point in their quest to set competitive standards. The organization conducted a smaller survey in 1988 with similar results.

“International assessment does give us a good picture of what being first would mean. It shows where we have to be. Until were know what we are shooting for, we are going to have a tough time being first in anything,” Anrig said.

Noting the survey’s findings that the top 10% of American students hold their own when matched with other countries’ top students, Anrig said, “We know how to teach math and science to our top kids. Our problem is how to get the others to do well too.”

Results of the survey, funded by the National Science Foundation, the U.S. Department of Education and the Carnegie Corp., were released in the face of growing criticism of international comparisons, which consistently find American students lacking.

Critics say the comparisons are invalid because many countries concentrate on educating--and testing--only their best students. For example, only half of China’s 13-year-olds are in school. Critics also say the tests can be misleading, lend themselves to superficial analysis and, most important, tend to divert attention from more meaningful school reform efforts.

“Different partisans that want to change the system use and abuse these findings that are not going to take us anywhere,” said Willis Hawley, director of the Institute for Public Policy Studies’ Center for Education and Human Development Policy at Vanderbilt University.

Hawley said there are difficulties in getting uniformly representative samples of students in various countries, and there is no way to account for the important, non-school-related factors that affect achievement. He cited differences in cultural values, the proportion of children living in poverty or in dysfunctional families, and in spending on various issues that affect children’s well-being, including health.

“It is pretty clear (from the survey) that our students can and should do better than they are, but the question becomes whether the international comparisons help us to improve.”

But many educators believe the comparisons help identify problems that too many Americans would otherwise ignore.

“No doubt there will be people who will find all sorts of technical errors in the way these exams were set up and administered and in the way the data are being read,” Albert Shanker, president of the American Federation of Teachers, said in a prepared statement.

“But when every report reaches the same conclusion, you have to suspect that they’re onto something,” said Shanker, who supports efforts to measure American students against their future competitors in the increasingly global economy.

Educational Testing Service officials said they heeded criticisms and took steps to make the assessments as fair as possible, including incorporating the National Academy of Sciences’ new standards and indicating countries in which test results include only a narrow segment of the two age groups.

Project Manager Archie E. Lapointe said one of the goals of the international assessment was to “look for recipes for success,” common threads that might be applied in overhauling the schools in countries where students fare poorly.

But survey officials found no clear answers. There seemed to be little correlation between achievement and such factors as class size, length of the school year and amounts spent on education. For example, the top-scoring South Koreans have the biggest class sizes--an average of 49 students in eighth grade.

“That leads us to conclude that each society has to find its own answer for success,” Lapointe said. “If I were (U.S. Education Secretary) Lamar Alexander, I would look at those schools that are teaching our top 10%.”

Comparing Test Scores

This chart shows how U.S. students performed on tests of math and science knowledge compared to students in 10 other countries. A total of 20 countries were surveyed by the Educational Testing Service last year. The numbers represent the percentage of questions answered correctly on the test.

COUNTRY STUDENT MATH PARTICIPATION Age 9 Age 13 Age 9 England Representative of 59 61 63 all students; low participation at both 9 and 13 Hungary Representative of 68 68 63 all students Israel Hebrew-speaking 64 63 61 schools Jordan Representative of -- 40 -- all students South Korea Representative of 75 73 68 all students Mozambique Cities of -- 28 -- Maputo and Beira; in-school population; low participation Scotland Representative of 66 61 62 all students Former Soviet Union 14 of 15 republics; 66 70 62 Russian-speaking schools; low participation at age 9 Switzerland 15 of 26 cantons -- 71 -- Taiwan Representative of 68 73 67 all students United States Representative of 58 55 65 all students

COUNTRY SCIENCE Age 13 England 69 Hungary 73 Israel 70 Jordan 57 South Korea 78 Mozambique -- Scotland 68 Former Soviet Union 71 Switzerland 74 Taiwan 76 United States 67

SOURCE: Educational Testing Service