Results from the recently released 2003 Trends in Mathematics and Science Study (TIMSS) and the Program for International Student Assessment (PISA), two major international studies of achievement in science and mathematics, provide insights and an opportunity to increase students’ science achievement. Many teachers of science at all levels involved in reform efforts are interested in knowing if their work has had an effect on our students’ ranking on these international tests.

Our scores

In 2003, TIMSS assessments were administered to carefully selected, equivalent samples of fourth- and eighth-grade students in 49 countries. The TIMSS tests have a curricular focus, with content items upon which all countries have agreed. The PISA study, sponsored by the Organization for Economic Cooperation and Development (OECD) and involving 32 countries, assessed the ability of 15-year-olds to apply science content and processes in life situations that affect individuals, communities, or the whole world.

The good news—results from the eighth-grade 2003 TIMSS tell us that U.S. student scores have improved significantly from 1995 to 2003—an increase of 12 points from 515 to 527. The United States posted the third-highest improvement in the 45 participating countries and was the only G-8 nation to show improvement. The United States is now the ninth-ranked country in average science achievement at the eighth-grade level. The countries outscoring the United States were the consistently high-performing Pacific Rim countries, as well as Hungary and Latvia.

Even better news was that the achievement gap in science between minorities and caucasian students at the eighth-grade level has decreased. African Americans showed remarkable and significant gains of 40 points, and Hispanic American achievement was up almost as much, with a 36-point gain above the 1995 scores. Although the achievement gap is still too large, these results are encouraging.

The news about our fourth graders is both good and bad. Fourth-grade U.S. scores in 2003 were essentially unchanged from those of 1995. (In the 1995 TIMSS, the United States ranked third in the sample of 44 nations, only behind Singapore and Japan; although a slight decline occurred in these scores since 1995, the achievement level was not significantly different.) At this level, the achievement gap for African Americans also narrowed, thanks to a 25-point increase in their scores. Unfortunately, the scores for Hispanic students were flat across the two testing periods, and fourth-grade caucasian students saw a drop in their scores from 1995.

At the high school level, the 2003 PISA study tells us that U.S. students did not fare well when compared to their international counterparts. In the previous PISA study conducted in 2000, there was no significant difference in the average science scores of U.S. students when compared to those of students in the participating OECD countries. However, this changed in the 2003 PISA assessment; U.S. students scored significantly lower than the international average and lower than 18 of the 32 countries participating in the study.

The analysis

So, what do these assessment results mean for the science education community? Before putting these international assessments in perspective, one must first consider the initiatives that are broad enough to affect change nationwide. Starting in about 1991, a series of systemic initiatives for states and urban and rural areas (and later local school districts) was funded by the National Science Foundation (NSF). In many areas these reform efforts continued well past 2003. The releases of the Benchmarks for Scientific Literacy in 1993 and the National Science Education Standards in 1996, followed shortly by standards in most of the states, provided another important boost to improving the quality and quantity of science teaching and achievement.

Is there evidence in the international studies that these efforts paid off? A small group of mathematics and science educators who convened at the Biological Sciences Curriculum Study headquarters in Colorado Springs to review the PISA and TIMSS results believe there is. Initiatives and funding for the improvement of science instruction during the past 10 to 12 years have been predominately directed at the elementary level, with only limited support provided to the middle level. Little support has been directed to high school science. One could argue that the 2003 TIMSS scores show that our eighth-grade students benefited from the accumulated funding and efforts of teachers and other science educators to implement standards, in addition to improved curriculum materials and instructional methods.

Although the already high scores of U.S. students at grade four in the 1995 study did not change significantly in TIMSS 2003, the data show that the achievement of fourth-grade caucasian students actually declined. This can be attributed to multiple and widespread reports from elementary teachers that the increased emphasis on literacy, which began in the early 1990s, squeezed much of the quality standards-based science out of the early elementary grades.

At the high school level, PISA results indicate that much more work and support is needed. Although the nature of the PISA addresses a slightly different domain of knowledge that emphasizes how well 15-year-olds can use and apply knowledge in personal and societal contexts, the fact is that U.S. students simply did not score well compared to their international peers. The data, coupled with the limited improvement efforts at the secondary level in the last 10 to 12 years, strongly suggest a need to provide the support that has been previously available at the elementary and middle school grades to students at the high school level. Science teachers should also teach students to apply science concepts and processes in personal and social contexts.

Moving forward

Making inferences from international data is challenging, but when assessment results match widespread reports from teachers in the field, the connection may be justified. Where patterns of support (or lack of it) for teachers and programs correspond with changes in student achievement scores over time, the speculation is worthy of further study. It could be used to guide the decisions of policy makers at national, state, and local levels who are interested in improving the quality of science achievement of all of our students.

Harold Pratt was the 2001–2002 NSTA President and is currently the President of Educational Consultants in Littleton, Colorado.