- Surveys & Programs
- Data & Tools
- Fast Facts
- News & Events
- Publications & Products
- About Us

Dr. Gary W. Phillips

Acting Commissioner of the National Center for Education Statistics

Pursuing Excellence: Comparisons of International Eighth-Grade Mathematics and Science Achievement from a U.S. Perspective, 1995 and 1999

December 5, 2000

Good morning. My name is Gary Phillips, the Acting Commissioner for the National Center for Education Statistics. Today the National Center is releasing the results of a repeat of the Third International Mathematics and Science Study (TIMSS-R). The United States' participation was also funded by our partner, the National Science Foundation. The study being released today was conducted in 1999 and is the successor to the 1995 Third International Mathematics and Science Study. Both studies were coordinated by the International Association for the Evaluation of Educational Achievement (IEA), a Netherlands-based organization dedicated to international comparative studies of education. The international reports were released in Boston this morning at 10:00 am. The report being released now is the United States report titled Pursuing Excellence.

Like its predecessor study, the 1999 survey responds to the U.S. education community's need for reliable and timely data on the mathematics and science achievement of U.S. students within an international context. This allows the United States to examine its education system through the prism of other countries' education systems to better understand different approaches to teaching and learning mathematics and science.

The data from TIMSS-R is actually much more comprehensive than the initial findings released today. The total study has four components.

- Mathematics and science assessments and student, teacher, and school questionnaires that ask for information on teaching practices, students' study habits, teacher training and professional development, and school policies. The initial results will be reported today.
- The second large component is a Benchmarking Project in which 27 states, districts, and consortia of districts throughout the United States voluntarily conducted the assessment in their own jurisdictions. The results the results to be released in April of next year.
- The third component is a Videotape Classroom Study that examines actual instructional practices in eighth-grade mathematics and science classes in 7 nations. These results will be released in October of next year.
- NAEP/TIMSS Linking Study that provides an opportunity to compare students' performance on the National Assessment of Educational Progress (NAEP) to their performance on TIMSS-R. This will be released in December of next year.

Taken together, these components provide a wealth of cross-national and national information on student performance and the context within which student learning takes place.

Although the 1995 TIMSS reported many interesting findings, possibly the most discussed finding was the declining standing of the United States among the 41 nations as children progress through school. The United States ranked above the international average in math and science in the fourth grade, dropped to the middle of the pack in the eighth grade, and by twelfth grade was among the lowest scoring nations in the world. This was true even of students who took physics and advanced mathematics. As you will see shortly, the 1999 TIMSS-R validated this finding.

The report being released today provides for three sets of international comparisons.

- In 1999, U.S. eighth graders exceeded the international average of 38 nations in mathematics and science. It should be noted that the set of 38 nations who participated in 1999 was different in important ways than the 41 nations who participated in the 1995 TIMSS. Several European countries (such as Switzerland, France, Austria, Ireland Germany, Spain and Portugal) did not participate, while many developing countries joined the study. At the same time, however, the highest scoring countries in 1995 also participated in 1999.
- Between 1995 and 1999, there was no change in eighth grade mathematics or science performance in the United States. In fact there was no improvement at the eighth grade in almost all of the 23 nations that participated 1995 and in 1999.
- The mathematics and science performance of the United States was lower for eighth graders in 1999 than it was for fourth graders four years earlier. This means that although U.S students learned a lot of math and science from the fourth to the eighth grade, the other nations learned more. This finding was obtained from the 17 nations that participated at fourth grade in TIMSS 1995 and at eighth-grade in TIMSS-R 1999.

Although it is a bit complicated to have three sets of countries in this report, each set tells its own interesting story.

That said, analysis of TIMSS-R data for the 38 participating nations indicates the following findings:

- U.S. eighth-grade students outperformed their peers in 17 nations, performed similarly to their peers in 6 nations, and performed lower than their peers in 14 nations in mathematics in 1999. In science, U.S. eighth-grade students outperformed their peers in 18 nations, performed similarly to their peers in 5 nations and performed lower than their peers in 14 nations in 1999.
- The report also reveals that U.S. eighth-grade students performed above the international average in 3 of the 5 mathematics content areas: fractions and number sense; data representation, analysis, and probability; and algebra; and U.S. eighth-grade students performed at the international average in measurement and geometry (figure 4, page 17 in the report). In science, U.S. eighth-grade students performed above the international average in 5 of the 6 content areas: earth science; life science; chemistry; environmental and resource issues; and scientific inquiry and the nature of science; and U.S. eighth-grade students performed at the international average in physics (figure 5, page 19 in the report).

Because TIMSS-R was specifically designed to allow for comparisons with results from TIMSS four years earlier, comparisons can be made between how our eighth-grade students fared in 1999 and how they fared in 1995. Twenty-three nations participated at the eighth-grade level in both TIMSS and TIMSS-R. This figure shows the average mathematics score for the 23 nations assessed in 1995 versus 1999.

This comparison reveals the following:

- Between 1995 and 1999 there was no change in eighth grade mathematics performance in the United States. Three nations-Canada, Cyprus, and Latvia-LSS-experienced increases in their mathematics average over the four years. The Czech Republic was the only nation to document a decrease in mathematics over the four years. The United States and 18 of the 23 nations showed no change in mathematics achievement at the eighth-grade level over the four years.
- The performance of the United States in mathematics over the four years is also reflected in mathematics content areas. In the 5 mathematics content areas in common between TIMSS and TIMSS-R, there was no change in the performance of U.S. eighth-graders or their peers in most of the other 22 nations. Only Canada and Latvia-LSS documented increases in performance in most of the 5 mathematics content areas.

Changes in average scores in science achievement among eighth-graders who took the TIMSS test in 1995 and those who took the TIMSS-R test in 1999 show similar results.

- When the nation's 1995 average score is compared to its own 1999 average score in science, the United States is one of 18 nations that experienced no change in the science achievement of its eighth-graders over the four years. Four nations experienced increases over the four years-Canada, Hungary, Latvia-LSS, and Lithuania; one nation experienced a decrease-Bulgaria.
- The performance of the United States in science over the four years is also reflected in science content areas. In the four science content areas in common between TIMSS and TIMSS-R, there was no change in the performance of U.S. eighth graders or their peers in most of the other 22 nations. Only Canada documented increases in performance in all four of the science content areas.

The next figure looks only at the subset of 17 nations whose fourth-graders participated in TIMSS 1995 and whose eighth-graders participated in TIMSS-R 1999.

- Data from TIMSS suggested that the international performance of the United States relative to other nations appeared lower at grade 8 in both mathematics and science than at grade 4. With the data available from TIMSS-R, it appears that the relative performance of the United States was lower for eighth-grade students in mathematics in 1999 than it was for this cohort four years earlier in 1995 for fourth-grade students. That is, U.S. fourth-graders performed at the international average of the 17 nations in 1995 while U.S. eighth-graders in this subset of nations performed below the international average in 1999.

In science, the story appears to be the same.

- In 1995, U.S. fourth-graders performed above the international average of the 17 nations in science, while U.S. eighth-graders performed at the international average in 1999.

The available evidence appears to confirm what had been suggested four years ago: that the relative performance of U.S. students in mathematics and science is lower at the eighth grade than at the fourth grade among this group of nations.

In terms of its performance on the assessments relative to other participating nations, the United States has not documented changes in mathematics or science between 1995 and 1999. However, as mentioned earlier, TIMSS-R includes a wealth of data on other areas of interest to researchers, policymakers, educators, and members of the public. In particular, TIMSS-R collected valuable information that can help place the achievement results into a broader context of the teaching and learning of mathematics and science both in and outside of school. Analyses of these data have identified a number of differences between the United States and the other participating TIMSS-R nations on such issues as teacher preparation, teacher attitudes, teaching practices, and classroom activities.

One topic that is of interest to policy makers is the academic preparation of the mathematics teachers of eighth-grade students. Over the last several years, some have argued that it is important for teachers to have subject matter expertise, and one indication of this is a major in subjects they teacher, either at the bachelor's or master's degree level. The mathematics teachers of the students assessed in TIMSS-R were asked to indicate their major or main area of study at the bachelor's or master's degree level. They could choose more than one area of concentration if applicable.

- There was a significant difference between the United States and the international average of the 38 participating nations on this characteristic. The results indicate that U.S. eighth-grade students were less likely than their international peers to be taught by a mathematics teacher with a bachelor's or master's degree in mathematics. In 1999, 41 percent of U.S. eighth-grade students had a mathematics teacher whose major or main area of study was in mathematics, a smaller percentage than the international average of 71 percent. U.S. eighth-grade students were as likely as their international peers to be taught mathematics by a teacher with a major in mathematics education (37 percent vs. 31 percent, respectively). Finally, U.S. eighth-grade students were more likely than their international peers to be taught mathematics by a teacher with a major in education (54 percent vs. 32 percent, respectively).

We collected similar information from science teachers. Science teachers often obtained degrees in different content areas of science such as biology, physics, and chemistry.

- U.S. eighth-grade students were as likely as their international peers to be taught science by a teacher with a bachelor's or master's degree major in biology, chemistry or science education. U.S. eighth-grade students were less likely than their international peers to be taught a science by teacher with a major in physics. Finally, as in mathematics, U.S. eighth-graders were more likely than their international peers to be taught science by a teacher with a major in education.

In addition to asking about the academic backgrounds of teachers, teachers were asked how confident they were to teach mathematics and science as a gauge of their own sense of preparedness.

- U.S. eighth-grade students were more likely than their international peers to be taught mathematics by a teacher who felt very well prepared to teach mathematics. In 1999, 90 percent of U.S. eighth-grade students were taught mathematics by teachers who felt very well prepared to teach mathematics. This compares to the international average of 73 percent.
- In science, U.S. eighth-grade students were also more likely than their international peers to be taught science by teachers who felt very well prepared to teach science. However, as the figure shows, a higher percentage of U.S. eighth-grade students were taught mathematics by teachers who felt very well prepared to teach mathematics than they were taught science by teachers who felt very well prepared to teach science. This is also the case internationally.

In addition to exploring the academic preparation and confidence levels of eighth-grade teachers, TIMSS-R collected information from teachers on the topics emphasized most in eighth-grade mathematics and science lessons.

- U.S. eighth-grade students were more likely than their international peers to be in classes where the teacher emphasized general mathematics or algebra as a single topic. A little more than half of U.S. eighth-graders were in a mathematics class where the emphasis was on general mathematics or algebra as a single topic; a similar percentage of their international peers were in classes where the emphasis was on a combination of algebra, geometry, and numbers. Another way of looking at the data indicates that 39 percent of U.S. eighth-graders in 1999 were in a class that emphasized geometry to some degree compared with 77 percent of their international peers.

In science in 1999,

- eighth-grade students in other nations were more likely than their U.S. peers to be in a class where the teacher emphasized general/integrated science or biology. On the other hand, U.S. eighth-graders were more likely than their international peers to be in a class where the emphasis was on earth science or physical science.

TIMSS-R also asked students and teachers about various practices and activities that took place in the classroom. The kinds of activities that students are asked to participate in during lessons can promote and reinforce learning, particularly when combined with a coherent and well-planned curriculum.

- Students were asked how often various activities occurred in their mathematics classes. A higher percentage of U.S. eighth-graders than their international peers reported that their teacher showed them how to do a mathematics problem or that they worked on worksheets or from the textbook almost always or pretty often in their mathematics classes. U.S. eighth-grade students reported that they were less likely than their international counterparts to work on mathematics projects almost always or pretty often in their mathematics classes.

The results for science paint a different picture

- While U.S. eighth-graders were more likely than their international peers to work on worksheets or from textbooks in their science classes, they were also more likely to work on science projects and conduct science experiments. U.S. eighth-grade students were as likely as their international peers to report that their teachers demonstrated a science experiment almost always or pretty often in their science classes.

There were also significant differences in the way in which U.S. students and their international peers were asked to deal with homework in their mathematics and science classes.

- U.S. eighth-graders were more likely than their international peers to discuss completed homework in their mathematics and science classes, but they were also more likely than their international peers to begin their homework in their mathematics or science classes. Seventy-four percent of U.S. eighth-graders reported that they almost always or pretty often began homework in their mathematics class in 1999 compared to 42 percent of their international peers. Fifty-seven percent of U.S. eighth-graders gave similar reports about beginning homework in their science classes compared to 41 percent of their international peers. These data seem to suggest that reviewing and working on homework assignments are a larger part of mathematics and science classroom activities in the U.S. than in other nations. Indeed, reviewing the prior day's homework and beginning the newly assigned homework during class in the United States was first noted in the 1995 TIMSS Videotape Classroom Study.

Comparisons of the achievement of selected population groups reflect a concern that all students-regardless of race, ethnicity -receive equitable educational opportunities. Has there been any change in the mathematics and science achievement of U.S. eighth-grade white, black, and Hispanic students over the four years?

- Like the overall national average in mathematics between 1995 and 1999, the performance of white and Hispanic eighth-graders in mathematics and science remained unchanged over the four years. However, eighth-grade black students improved their performance in mathematics over the four years. They, along with the other groups, did not show an improvement in their science performance, however.

There is a great deal of other useful information in the report released today by NCES. Briefly, the other highlights are:

- In both 1995 and in 1999, there was no difference in mathematics achievement between U.S. eighth-grade boys and girls;
- In both 1995 and in 1999, U.S. eighth-grade boys outperformed U.S. eighth-grade girls in science;
- U.S. eighth-grade students whose mother or father attended some college or completed college showed an increase in performance in mathematics between 1995 and 1999; and
- U.S. eighth-grade students whose mother or father completed college showed an increase in performance in science between 1995 and 1999.
- Both U.S. eighth-grade public school and nonpublic school students performed higher than the international average in mathematics and science in 1999. Furthermore, the report indicates that U.S. eighth-grade nonpublic school students outperformed their public school counterparts in both mathematics and science in 1999.

In conclusion, there are four points to be made. First, TIMSS-R includes a wealth of information on mathematics and science teaching and learning in 38 nations. The achievement and questionnaire results presented today are only the first findings to be reported from TIMSS-R. Over the next couple years, the Center will continue to provide additional information by reporting on in-depth analyses of achievement data and releasing the findings from the Benchmarking study, the Videotape Classroom Study and the NAEP/TIMSS-R Linking Study.

Second, TIMSS-R is one of several studies of mathematics and science achievement of U.S. students that NCES conducts. In addition to TIMSS-R, NCES conducts the National Assessment of Educational Progress (NAEP) and participates in the Program for International Student Assessment (PISA), an international study of the reading, mathematics, and science literacy of 15-year old students in nations that are members of the Organisation for Economic Cooperation Development. Over the next year, NCES will be releasing additional findings from TIMSS-R, as well as from NAEP and PISA. The Center invites everyone to utilize the information provided by each of these projects to the fullest extent possible in efforts to improve the mathematics and science education of all students.

Third, the IEA-the organization that oversees the TIMSS project-has announced its intention to conduct a next TIMSS in 2003. Before making a decision whether or not to commit to participating in TIMSS 2003, NCES and NSF will carefully discuss the opportunity.

Lastly, the effort to carry out and report on TIMSS-R in the United States is the result of a major collaborative effort between NCES and NSF, with support from the Secretary and the assistance of the U.S. TIMSS-R Technical Review Panel. It is the intention of the Center to continue its work with NSF in this endeavor, and to provide researchers, educators, policymakers, and the public with the most up-to-date and relevant information possible.

We would also like to thank the principals, teachers, and students who participated in this study. The positive response of educators to this international study testifies to its value.

Let me also take a moment to acknowledge the authors of this report. I know that they have worked long and hard to get to this day. Authors include: Patrick Gonzales, Christopher Calsyn, Leslie Jocelyn, Kitty Mak, David Kastberg, Sousan Arafeh, Trevor Williams, and Winnie Tsen. Thank you again for your hard work.

Now that the results have been released, I would like to turn to our distinguished guests today. First, Secretary of Education Richard Riley: Secretary Riley is a strong supporter of providing the best possible education to American children, and his support extends to the collection of data like TIMSS-R, which is needed to take measure of the progress that the American educational system has made in improving education for all students.

Next is the Director of the National Science Foundation, Dr. Rita Colwell. Prior to becoming Director at NSF, Dr. Colwell has held numerous advisory positions to the government and private foundations. She is a nationally-respected scientist and educator. NSF is a co-funder of TIMSS-R and a key partner in this endeavor, and the Center has very much appreciated Dr. Colwell's strong support.

Finally, I want to introduce the co-chairs of the U.S. TIMSS-R Technical Review Panel, Dr. Susan Fuhrman and Dr. Margaret Cozzens: Dr. Fuhrman is Dean of the Graduate School of Education at the University of Pennsylvania and chair of the management committee of the Consortium for Policy Research in Education (CPRE), a joint research project on state and local education policies and finance. Her work focuses on education policy and finance as well as state education reform, and state-local relationships, among other areas. Dr. Margaret Cozzens is Vice Chancellor for Academic and Student Affairs at the University of Colorado at Denver and professor of mathematics. Dr. Cozzens' work focuses on the improvement of education for all students (preK-graduate school), especially in the areas of mathematics and science, and teacher professional development. Through their leadership of the Technical Review Panel, they and the other members of the TRP have provided critical advice and support of the TIMSS-R project.