How do we know this? Our 12th graders along with students in their last year of school in more than 20 countries were tested in 1995 as part of the Third International Mathematics and Science Study, better known as TIMSS. The results are included in this report, Pursuing Excellence: A Study of U.S. Twelfth-Grade Mathematics and Science Achievement in International Context. This is the third report from the study. Previous reports covered international comparisons of 4th graders and 8th graders.

In sheer quantitative terms, TIMSS is the world's largest, most comprehensive, and most rigorous international education comparison ever. In 1995, researchers tested the mathematics and science knowledge of more than half a million students in 41 countries at three grade levels-primary, middle, and end of secondary. This is far more than any previous study. TIMSS ensured that the participating students in each country were representative of its population -- this study is a fair evaluation.

Today we are here to present the results of TIMSS covering the end of secondary school. We actually gave four different tests to these students. In 21 countries, students were tested for general knowledge in mathematics and science. In addition, in 16 countries, advanced students were tested in physics and advanced mathematics. This study included largely the developed European countries, plus the U.S., Canada, Australia, and New Zealand. Asian countries chose not to participate.

TIMSS is not an assessment of other country's best students against our average students, but of the entire range of students in each country. While the percentage of young adults who complete secondary school in the U.S. once was significantly larger than the percentage in other countries, this is no longer the case. Today, similar proportions of young people are enrolled at the end of secondary schooling in most of these countries. Since some nations group their students into different types of schools with different graduation requirements, TIMSS assessed students in their last year in all types of schools and programs in all countries. This last grade ranged from the ninth grade in some vocational programs through the 14th grade in other programs.

Results of the General Knowledge Assessments
When we look at the results, we see that the U.S. was among the lowest performing countries on both the mathematics and science general knowledge assessments. U.S. performance was below the international average in both mathematics and science. In mathematics we were outperformed by 14 out of 20 countries, were similar to four countries, and outperformed two countries. In science we were outperformed by 11 countries, were similar to seven, and outperformed the same two countries. This relatively low U.S. performance is not a change from patterns of previous international assessments at this grade level. It also continues the pattern of slightly better performance in science than in mathematics.

Please note that in talking about the U.S. performance relative to other countries, we should not try to rank them in strict numerical order. Because the data are based on samples of students each score has a statistical margin of error and really represents a range of possible scores. For this reason, we refer to bands of performance for countries that perform better than, similar to, or poorer than the U.S. It would be statistically wrong to say that one country was better than the U.S. if it is in the same group, even if its average score looks different than ours.

Contrary to myths about U.S. education, our poor performance is not because our student body is more diverse or because we have a lot of low scores pulling down the overall U.S. average. Most countries have students from diverse language and cultural groups. It is also true that the other countries in this study have a similar range of performance as the U.S., covering 300 points from the 5th to 95th percentile. However, the entire distribution of U.S. scores both starts and ends lower than in most other nations. This means that the average level of general knowledge in mathematics among students in a majority of these countries matched that of the top quarter of U.S. students. Similarly, a student scoring at the 50th percentile in mathematics in the U.S. would be at about the 25th percentile (or below) in 12 nations. Therefore, we cannot blame our overall low performance on our bottom students; the problem is across the board.

When we compare these 12th grade TIMSS results to the previously released fourth and eighth grade results, we see a loss of competitive advantage. This means that, when comparing the same groups of countries across grades, the relative standing of U.S. eighth graders is lower than that of our fourth graders; and the standing of our 12th graders is lower than that of our eighth graders.

Results of the Advanced Students Assessments
U.S. student performance on the assessments in Advanced Mathematics and Physics was among the lowest of participating countries and, in both cases, below the international average. On the Advanced Mathematics assessment, the U.S. was outperformed by 11 countries, was similar to four countries, and did not outperform any country. On the Physics assessment, the U.S. was outperformed by 14 countries, was the same as one country, and did not outperform any country. On all five content areas that make up Physics and all three content areas comprising Advanced Mathematics, the U.S. was below the international average and was outperformed by a majority of the other countries.

Who took the advanced TIMSS assessments? A random sample of the approximately 10-20 percent of students in the last year of school in each country took these tests. In the U.S. the students who took the Advanced Mathematics TIMSS had already taken or were currently enrolled in precalculus, calculus, or Advanced Placement Calculus. They represented 14 percent of young people their age in the U.S. compared with the 19 percent average for the 16 countries as a whole. In the U.S. the students who took the Physics assessment had previously taken or were currently enrolled in Physics. They represented 14 percent of young people their age which was the same as the international average for all participating countries. Therefore, the U.S. performance was not the result of any differences in the selectivity of students taking the assessments.

When we compare these international students to our most advanced students, we come closer to the international average. In the U.S., many of our advanced students do not take calculus, but a quarter of the items on TIMSS Advanced Mathematics assessment are Calculus questions. When we looked at U.S. students who had taken calculus, we found that their performance were close to the international advanced student average. Six countries outperformed this U.S. Calculus population, seven were similar, and the U.S. outperformed two countries. If we compare only those U.S. students with Advanced Placement Calculus to the international advanced mathematics group, only France outscored the U.S. Advanced Placement students. The U.S. outperformed five countries and seven were similar to the U.S. In Physics, when the U.S. Advanced Placement Physics students were compared with the Physics students in other countries, the U.S. students were outperformed by four countries, were similar to ten countries, and outperformed one country.

Factors Related to Performance
Our analysis of TIMSS data do not suggest any single cause of this level of U.S. performance. For example:

• Television: Students in other countries watch just as much television as our students;
  
• Part-time jobs: More U.S. students work in part-time jobs and work more hours than students in countries that scored higher and lower than us; and
  
• Homework: Students in some countries that outperform us do less homework and studying than we do, while students in other countries that we outperform do more.

Conclusion
TIMSS provides good information to help educators, policymakers, and the public make well-informed decisions. Sound data support sound policies. And TIMSS reports are just a starting point. The TIMSS resource kit, which includes materials on curriculum, teaching, and achievement, helps educators to evaluate their own schools.

The National Center for Education Statistics (NCES) is committed to learning more from TIMSS. We will continue to make this rich data base widely available and to sponsor further research on the relationship between student achievement and educational structures, processes, and student experiences around the world.

We will be conducting a second round of TIMSS. I am pleased to announce that in the Spring of 1999, my Center will collaborate with the National Science Foundation and our international partners on an international assessment to determine whether the fourth graders of 1995 who performed so well on TIMSS are able to maintain their high comparative standing four years later as eighth graders. This will provide valuable insight on the effects of the education reforms of the 1990s.

In the United States, we call graduation commencement. It is not just an ending to one's school career, but also a beginning of the next phase in one's life. I see today's unveiling of the TIMSS results in the same way. It is not an ending, a final calculation of our students' achievement. It is really the beginning of our efforts to use this valuable information to improve our education system and to raise the performance of our students.