Copyright © 1998 by The New York Times Co. Reprinted by Permission.

The dismal performance of American 12th graders on the Third International Mathematics and Science Study is naturally disheartening. Even in a pool that lacked the usual high-scoring Asian countries, American students managed to score near the bottom.

But we should resist the tendency to focus on increasing our students’ scores on these tests. These tests don’t measure whether students can think scientifically or mathematically, they just measure a kind of lowest common denominator of facts and skills. So getting students to do well on them doesn’t necessarily mean much in the real world. It doesn’t even mean that students will have successful careers in science and technology.

Half a dozen years ago, when our economy was languishing along with our test scores, it was easy to blame our poor schools and to push for better results from our students. Now the United States stands at the top of the world economically, but our students are still scoring at the bottom on international math and science tests. Since high scores on these tests obviously aren’t crucial to our economic success, we need to decide what kinds of tests matter in helping form the kind of citizen we want to have.

Standardized tests should find out if American students can think in a scientific manner. That means teaching students to understand the nature of the scientific method: how experiments are set up, how models and theories are constructed and tested, how to decide what theories best describe a phenomenon. Students learn these concepts only by designing and carrying out their own experiments.

Tests should measure a student’s scientific ability by presenting a problem, offering data to solve that problem and positing several different interpretations of the data. The test taker might be asked to determine what conclusions can and cannot be drawn.

But most standardized tests ask fact-based questions that sample a wide range of topics in a somewhat superficial way. Students who score well on these tests are like well-trained athletes or musicians: through practice, they have become proficient at a certain skill — in this case, they have done extensive problem sets in many different “content areas” and can move quickly from one question to another.

These tests are helpful in the real world, especially in high school and college. But they simply do not show whether a student can think seriously about a scientific issue. We could drill our students with problem sets and raise their test scores, but still be left with a population that remains scientifically and mathematically challenged.

After all, students should be able to apply scientific and mathematical concepts to the world around them. As adults, they need to know how to decide which life insurance to buy, how pesticides affect their food and how interest rates determine home mortgages. Citizens also need to be able to decide whether cloning research should be banned, whether more money should be poured into studying global warming and whether there should be a national health-care plan.

Aren’t these skills ultimately important for our nation and others, too? We must rethink the ways we teach science and math. We must have the foresight to pursue a less-traveled road, one that can lead to a citizenry that can think rigorously and make informed decisions and that can handle a future where science and technology will be more important than ever.