U.S. Students' Scores Drop By 8th Grade
In 1995, the nation's 4th graders aced international mathematics and science tests. By the time they reached the 8th grade in 1999, though, they had become little better than C students on a global curve, a study released here last week concludes.
For More Information
|The report "Pursuing Excellence: Comparisons of International Eighth-Grade Mathematics and Science Achievement from a U.S. Perspective," is available from the National Center for Education Statistics. (Requires Adobe's Acrobat Reader.)|
The scores from the 38-nation testing project reinforce what many education researchers have said for a long time: American schoolchildren fail to sustain the achievement advantages they gain in elementary school.
"Our curriculum includes a lot of repetition and fewer new subjects than in other countries," said Susan H. Fuhrman, the dean of the University of Pennsylvania's graduate school of education and a co-chairwoman of the panel of scholars that oversaw U.S. participation in the Third International Mathematics and Science Study—Repeat.
"We're not challenging our students to the extent that we can," Ms. Fuhrman added at the news conference held here to release the report.
For the so-called TIMSS-R, the International Association for the Evaluation of Educational Achievement, the study's Amsterdam, Netherlands-based sponsor, gave the same battery of exams to 8th graders that had been administered in 1995. Fourth and 12th graders were not tested this time around, as they were in 1995.
In 1999, American 8th graders scored above the 38-nation average both in science and math, but failed to distinguish themselves as high achievers in either subject. U.S. scores ranked in the same class as Bulgaria, Latvia, and New Zealand, among other countries. They fell well behind top scorers in Asia and Europe.
The highest-scoring countries in both math and science included Singapore, Japan, South Korea, Japan, and the Netherlands. When comparing 8th grade scores in 1995 with those last year, only Canadian and Latvian students made statistically significant gains in both subjects. Most countries' scores did not differ widely enough to be considered significantly changed.
Between 2,300 and 9,000 8th graders in each country took the test, depending on the size and demographic diversity of the nation. The United States gave the test to 9,000 students in 221 schools. On the other end of the scale, Lithuania administered the test to 2,300 children.
While the overall picture of 8th grade performance did not change much between 1995 and 1999, some experts voiced optimism that the achievement gap between minority and white students in the United States on the tests has narrowed. The initial signs of the shrinking gap are subject to further analysis, said Margaret B. Cozzens, the vice chancellor of academic and student affairs at the University of Colorado and a co-chairwoman of the U.S. TIMSS-R Technical Review Panel.
If the achievement gap is starting to close, it would reverse a decade-long trend toward a greater disparity. On tests such as the National Assessment of Educational Progress, however, the gap continues to expand.
For many scholars, though, the bigger story was American students' drop in standing between the 4th and 8th grades. In 1995, U.S. 4th graders bested the international average in both mathematics and science, ranking in the upper tier of science achievement and near the top in math in the 29-nation survey. ("4th Graders Do Well in Math, Science Study," June 18, 1997.) By the time the same group of students reached the 8th grade, its performance was mediocre.
When U.S. 4th graders scored better than 8th and 12th graders on the 1995 tests, many speculated that curriculum changes made in the early 1990s were paying off for the nation's youngest students. Some policymakers and educators predicted that the achievement would be sustained as the students got older.
But the TIMSS-R results undercut that theory.
"Some people thought the 4th graders [in 1995] reflected a new trend," said William H. Schmidt, a professor of education at Michigan State University and the head of a project that compared the curricula of all countries that participated in the 1995 TIMSS project. "This says that's not true. We don't have a new cohort of high-performing kids."
Only 17 nations had 4th graders take TIMSS in 1995 and 8th graders retake it four years later. In 1995, U.S. 4th graders scored at the average of that smaller group in math. By the time those 4th graders were in 8th grade last year, their achievement had dropped to the bottom tier, ahead of only five other countries.
In science, the 4th graders scored well above the 17-country average in 1995, but slipped below it four years later.
The data suggest specific places where American students lag. In mathematics, they perform below the international average in geometry and measurement, but score well in fractions and algebra. In science, they score the lowest on questions dealing with physics, chemistry, and earth science, but show well on environmental issues.
In his research for the 1995 TIMSS, Mr. Schmidt found a link between what is typically taught and how well the students performed on the tests.
In the United States, math and science curricula fail to introduce gradually new and more challenging topics and concepts, according to Mr. Schmidt. Instead, the curriculum typically reinforces skills learned in earlier grades. In 7th and 8th grades, most U.S. students are still reviewing arithmetic that children in other countries have mastered by the 5th grade, he said, and have moved on to study geometry and algebra.
The results may suggest that students in the United States spend so much time practicing specific skills that they never understand the mathematical principles that underlie them.
In companion surveys of teachers and students, TIMSS-R researchers found that U.S. teachers, especially in mathematics, focus on drill and practice. Ninety-four percent of students said their teachers "almost always" or "pretty often" showed them how to solve math problems when they were stuck. That compares with 86 percent of the total in the 38-nation survey.
Similarly, 79 percent of the American students said they "almost always" or "pretty often" were given time to do homework in class, while only 55 percent in the overall survey said that.
"The practices of demonstrating how to do math problems and of setting students to work on worksheets and textbook exercises were still much more common—and nearly universal—in the United States than they were in other countries," Ms. Cozzens said at the press conference.
Students abroad are more likely to be engaged in projects where they are discovering mathematical or scientific concepts and applying them in real-life situations, she added.
Another possible factor in the drop-off in U.S. performance is that teachers themselves haven't mastered the subjects they teach, Ms. Cozzens and Ms. Fuhrman suggested.
Only 41 percent of U.S. 8th graders are taught by math teachers who have math degrees, compared with 71 percent in the international survey. The disparity was not as pronounced in science, the TIMSS-R surveys found.
"Students deserve teachers who have a deep and conceptual understanding of specific content areas," Ms. Fuhrman said. "TIMSS-R reminds us that this is an area in which there is progress still to be made."
Mr. Schmidt said in an interview that American schools need to address the quality of what is taught and of who teaches.
"If you put in place a new curriculum, what stands right next to that is: Do you have the teachers with the subject-matter knowledge to pull it off?" he said.
Watching for a Payoff
The TIMSS-R results are no surprise to those who advocate changing the way schools teach math and science.
Any effort to make significant changes takes longer than the 21/2 years between the time the first round of TIMSS results were released in late 1996 and the time 8th graders took the test again last year.
In southwestern Pennsylvania, a privately subsidized effort to overhaul math and science teaching in 100 districts is only now getting new curricula in place.
It took one year just to redesign the curricula, a year to pilot-test them, and another to get all teachers using them, according to Nancy R. Bunt, the executive director of the Collaboratives for Learning, a Pittsburgh-based nonprofit group.
"Four years is a very short time in terms of showing results," she said. "If you're going to change the way you're teaching, it's not going to happen overnight."
But the group may know if the early seeds of its work are paying off. Schools in 11 counties in the Pittsburgh area gave the exams to a large enough sample of students to allow their test scores to be compared with those of children in other countries.
Vol. 20, Issue 15, Pages 1,20