TIMSS: Why U.S. Students Performed Poorly

Published February 1, 1999

Results from the Third International Mathematics and Science Study (TIMSS) show that the longer U.S. students remain in school, the lower their world standing in math and science: Fourth-graders scored among the best in the world, eighth-graders were merely mediocre, and twelfth-graders placed close to last. But critics now argue that TIMSS methods are bogus, while apologists claim that low scores in a booming economy are no disgrace. Who’s right?

The Thomas B. Fordham Foundation has produced A TIMSS Primer: Lessons and Implications for U.S. Education, written by University of Michigan psychology professor Harold W. Stevenson, a leading authority on international comparisons of K-12 education. As director of the TIMSS case study project, Stevenson developed the study methodology and produced its findings.

Although the major news focus has been on TIMSS achievement tests, there are four other components to TIMSS: analyses of academic standards and curricula in different countries; surveys of students, teachers, and administrators, including study habits; in-depth subject analysis, including national standards and teacher training; and analyses of videotaped classroom lessons.

Among the possible explanations Stevenson offers for U.S. students’ poor TIMSS performance: a fragmented curricula; an emphasis on developing rules to apply rather than understanding why; a belief that family background rather than hard work determines academic success; a lack of clear, tough academic standards; and the low status awarded teachers in U.S. culture.

For more information …

Harold W. Stevenson’s July 1998 report, A TIMSS Primer: Lessons and Implications for U.S. Education, was published by the Thomas B. Fordham Foundation, 1015 18th Street NW, Washington, DC 20036, 202/223-5452. Copies are available from the Foundation by calling 1-888-TBF-7474. The report also is available through PolicyBot. Point your Web browser to http://www.heartland.org, click on the PolicyBot icon, request old documents #2136412 (part 1, 19 pp.) and #2136413 (part 2, 20 pp.).