Today the National Center for Education Statistics is releasing results on the performance of U.S. students from the Trends in International Mathematics and Science Study (TIMSS). TIMSS, conducted every four years, is an assessment of fourth- and eighth-graders in mathematics and science. TIMSS first collected data in 1995, and then again from eighth-graders in 1999. With the 2003 data collection, TIMSS offers the first international trend comparisons in mathematics and science at grades four and eight.

In 2003, the United States and a number of other countries participated in data collection at two grade levels: 25 nations collected data on fourth-graders and 45 nations collected data on eighth-graders. For 15 of these nations, including the United States, TIMSS offers comparisons of fourth-grade student achievement between 1995 and 2003. For 34 of these nations, including the United States, TIMSS also offers comparisons of eighth-grade student achievement between 2003 and at least one prior data collection year, either 1995 or 1999.

Features of the Assessment
Content
TIMSS measures how well students acquired the mathematics and science knowledge that they have encountered in school. That is, the content of the TIMSS assessment is based on the curricula of participating countries. Like the National Assessment of Educational Progress (NAEP), TIMSS assessments are based on collaboratively developed frameworks on the topics from curricula in mathematics and science that should be assessed; but unlike NAEP, the framework and consensus process involves content experts, education professionals and measurement specialists from many different countries, rather than just the United States.

Because countries vary in the ways in which mathematics and science is taught as well as the content covered in their school-based courses, this means that the TIMSS assessment should be considered a general indicator of the knowledge of a nation’s students. For example, in some countries, large proportions of eighth-graders have been exposed to early and advanced topics in algebra and geometry, whereas in the United States, a significant proportion of eighth-graders have not yet been exposed to these topics, or have only encountered the earliest notions.

In comparison to NAEP and the Program for International Student Assessment (PISA), another international assessment for which results were released one week ago, TIMSS and NAEP share the most in common, in terms of content and cognitive demand, particularly in mathematics. As with any assessment, it is important to consider the TIMSS results in relation to other sources of data, such as NAEP and PISA.

Sponsorship
TIMSS is sponsored internationally by the International Association for the Evaluation of Educational Achievement (IEA), an organization of national research institutions and governmental research agencies. The IEA is releasing the international report today, while the NCES report focuses on U.S. performance.

How TIMSS Was Conducted
The samples in each TIMSS participating country were selected to be representative of all students in the upper of the two grades containing the most number of 9- and 13-year olds. This corresponds to the fourth and eighth grade in nearly all of the countries, including the United States. The U.S. sample included public and private schools in the United States. A total of 9,829 students in 248 schools participated at grade four, and 8,912 students in 232 schools participated at grade eight in 2003.

The TIMSS Scale
TIMSS results are reported on a scale from 0 to 1,000, with the international standard deviation set at 100. The majority of student scores fall between 200 and 800, however. The mathematics and science items were scaled separately. Thus, direct comparisons between the two scales are not advisable, nor can or should the TIMSS scores be directly compared to scores from other assessments.

Standard deviations associated with the mathematics and science scores in each participating country are included in the full report, in appendix C.

U.S. Performance in Mathematics
Performance at Grade Four
In 2003, U.S. fourth-grade students scored 518 in mathematics, on average, exceeding the international average of 495 for the 25 participating countries. U.S. fourth-graders were outperformed by their peers in 11 countries, including four Asian countries (Chinese Taipei, Hong Kong SAR, Japan, and Singapore) and seven European countries (Belgium-Flemish, England, Hungary, Latvia, Lithuania, Netherlands, and the Russian Federation). On the other hand, U.S. fourth-graders outscored students in 13 countries.

The United States is one of 15 countries that also collected data on fourth-graders in 1995. In 1995, U.S. fourth-graders also scored 518 in mathematics, on average, indicating that there has been no change in the average mathematics performance of U.S. fourth-graders over these 8 years. Of the 11 countries in which fourth-graders outperformed U.S. fourth-graders in 2003, fourth-graders in three countries demonstrated improvement in average mathematics performance over the same period (England, Latvia-LSS, and Hong Kong SAR).

Although there was no change in the average mathematics performance of U.S. fourth-graders between 1995 and 2003, their performance relative to their peers in the other 14 countries that also collected data at these two points in time was lower in 2003 than in 1995. While in 1995, U.S. fourth-graders were outperformed by their peers in four countries and performed better than their peers in nine countries, in 2003 U.S. fourth-graders were outperformed by their peers in seven countries, two of which experienced double-digit gains.

Performance at Grade Eight
In 2003, U.S. eighth-grade students scored 504 in mathematics, on average, exceeding the international average of 466 for the 45 participating countries. U.S. eighth-graders were outperformed by their peers in nine countries, including five Asian countries (Chinese Taipei, Hong Kong SAR, Japan, Korea, and Singapore) and four European countries (Belgium-Flemish, Estonia, Hungary, and the Netherlands). On the other hand, U.S. eighth-graders outscored students in 25 countries.

In contrast to their fourth-grade counterparts, U.S. eighth-graders improved in mathematics between 1995 and 2003, from an average score of 492 in 1995 to an average of 504 in 2003. There was no measurable change detected when comparing the average mathematics score of U.S. eighth-graders in the more recent period, from1999 and 2003. As a consequence of the improvement in the average mathematics performance of U.S. eighth-graders between 1995 and 2003 and the decline in performance over these 8 years in several countries in which eighth-graders had outperformed U.S. students in 1995, the relative performance of U.S. eighth-graders in mathematics was higher in 2003 than in 1995.

U.S. Subpopulation Performance in Mathematics
Performance at Grade Four
As with other assessments, TIMSS also provides information on the average performance of student subpopulations. Among those groups examined at grade four are boys and girls, race/ethnicity, and students who attend public schools with varying levels of eligibility for the federal free or reduced price lunch program, a proxy measure for poverty.

No measurable change was detected in the average mathematics achievement of U.S. fourth-grade boys or girls between 1995 and 2003. Nonetheless, U.S. boys outperformed girls in mathematics in 2003, which differs from 1995 when there was no measurable difference.

Black fourth-grade students in the United States demonstrated an improvement in average mathematics achievement between 1995 and 2003; they were the only racial/ethnic group to show measurable change in mathematics. Moreover, as a result, the gap in average scores between White and Black fourth-grade students in the United States narrowed over these 8 years, from 84 score points in 1995 to 69 score points in 2003.