Reassessing U.S. International Mathematics Performance: New Findings from the 2003 TIMSS and PISA
January 04, 2006
American Institutes for Research
As reports go, this one is as close to a thriller read as it gets. In this brief but fascinating analysis American Institutes for Research (AIR) brings fresh, and surprising, insights to the 2003 Trends in International Mathematics and Science Study (TIMSS) results for students in grades 4 and 8, and the 2003 Program for International Student Assessment (PISA) results for students aged 15. Many folks have used these results to demonstrate our schools' mediocrity relative to those in other nations, especially at the secondary school level. For example, our own Mayhem in the Middle report noted that "in math, our fourth graders were at the international average, but by eighth grade, students in 27 other countries scored higher than U.S. students, with significant differences in 20 of those." (These numbers are based on 1995 TIMSS scores.) While AIR doesn't dispute such points (in fact, it finds "consistently mediocre U.S. results on all three assessments"), it does say the reality is not so simple. To begin, the three assessments (TIMSS-4, TIMSS-8, and PISA) tested different countries. When one examines our relative standing against a common subset of those countries, we fare the same across all three assessments - U.S. scores do not plummet as the children age (we're consistently ranked 8th or 9th out of 12 nations). But this is just the beginning. Noting that TIMSS focuses on content knowledge, while PISA emphasizes real-world applications, the AIR study debunks the myth that math class should increasingly emphasize the "real-world" to correct our PISA failings. Since U.S. students don't do too well on either TIMSS or PISA, the nation's math shortcomings are hardly limited to just one area. The study also explores gender differences, showing that only in the U.S. and Italy do boys outperform girls in math in grades 4 and 8, while by age 15 boys score better across all countries. And, it finds no evidence that the female drop-off is caused, as some suggest, by girls' weakening confidence in their own abilities. Moreover, the study shows that "statistics is the area of clear U.S. strength," while we do "relatively poorly on measurement and geometry." AIR argues for a national curriculum, noting that without one the U.S. has absurdly broad textbooks - as publishers strive to cover the topics prescribed by 50 different state standards. Consequently, we sacrifice instructional time on the essential topics. (Total teaching time is not holding the U.S. back - we're about average. We just allocate our time differently.) And finally, not surprisingly, they point out that we have far fewer math teachers with degrees in math than do other countries. It's an impressive amount of useful data and conclusions packed into a mere 32 pages. For your own copy, click here.