February 2, 2015

U.S. students improving – slowly – in math and science, but still lagging internationally

Scientists and the general public have markedly different views on any number of topics, from evolution to climate change to genetically modified foods. But one thing both groups agree on is that science and math education in the U.S. leaves much to be desired.

In a new Pew Research Center report, only 29% of Americans rated their country’s K-12 education in science, technology, engineering and mathematics (known as STEM) as above average or the best in the world. Scientists were even more critical: A companion survey of members of the American Association for the Advancement of Science found that just 16% called U.S. K-12 STEM education the best or above average; 46%, in contrast, said K-12 STEM in the U.S. was below average.

Standardized test results appear to largely bear out those perceptions. While U.S. students are scoring higher on national math assessments than they did two decades ago (data from science tests are sketchier), they still rank around the middle of the pack in international comparisons, and behind many other advanced industrial nations.

STEM_pisaOne of the biggest cross-national tests is the Program for International Student Assessment (PISA), which every three years measures reading ability, math and science literacy and other key skills among 15-year-olds in dozens of developed and developing countries. The most recent PISA results, from 2012, placed the U.S. an unimpressive 35th out of 64 countries in math and 27th in science. Among the 34 members of the Organization for Economic Cooperation and Development, which sponsors the PISA initiative, the U.S. ranked 27th in math and 20th in science.

Younger American students fare somewhat better on a similar cross-national assessment, the Trends in International Mathematics and Science Study. That study, known as TIMSS, has tested students in grades four and eight every four years since 1995. In the most recent tests, from 2011, seven countries (out of 50 total) had statistically higher average fourth-grade math scores than the U.S., while six countries had higher average science scores. In the eighth-grade tests, six out of 42 countries had statistically higher average math scores than the U.S., and eight had higher science scores.

STEM_mathProfAnother long-running testing effort, the National Assessment of Educational Progress (a project of the federal Education Department), has found that U.S. students have made substantial math gains since 1990. A National Science Foundation report notes that, while eighth-grade scores “show a continuous upward trend, fourth-grade scores leveled off in recent years.”

The average fourth-grade NAEP math score in 2013 was 242 (on a 0-to-500 scale), versus 213 in 1990, but has moved up only 2 points since 2007. The average eighth-grade score was 285 in 2013, compared with 263 in 1990; it’s moved up 4 points since 2007.

Looked at another way, the 2013 NAEP rated 42% of fourth-graders and 36% of eighth-graders as “proficient” or “advanced” in math. While far fewer students now rate at the lowest performance level (17% of fourth-graders and 26% of eighth-graders, versus 50% and 48%, respectively, in 1990), improvement in the top levels has slowed considerably since 2007.

NAEP also tests U.S. students on science, though not as regularly. The 2013 results haven’t yet been released, and a major revision in 2009 means results before then aren’t comparable. Still, we can say that science performance among eighth-graders improved slightly between 2009 and 2011, with the average score rising from 150 to 152 (on a 0-to-300 scale). In 2011, just 32% were rated proficient or better, while a third were rated “basic” (indicating partial mastery of the material) and 35% were below basic.

Topics: Education, Educational Attainment, Science and Innovation

  1. Photo of Drew DeSilver

    is a senior writer at Pew Research Center.


  1. Anonymous10 months ago

    PISA doesn’t just test all the students in one country and then only the smarter ones in another, they break down the results based on the levels. The US still fairs very poorly when you look at their Top Percentile Students compared to other Top Percentile Students, it’s not stupid vs. brainiacs with the results.

  2. Anonymous10 months ago

    If the United States’ score is from the TOP five or ten percent, then there is definitely a huge problem here.

  3. MLG M81 year ago

    W8 m8 the united st8s were b8ed into doing math and sience cuz the asins are have pro strategies 5 test sniping and are 3 clever 5 us so we got rekt

    1. Anonymous10 months ago

      I’m just going to go ahead and say that the reason we have such poor test scores in America is because of people who have an education level like MLG M8 here.

  4. Jeff Landrum1 year ago

    On the countries that scored above the U.S., what percentage of total students were tested?
    My district prides itself in testing every student, not the upper 5 or 10% which I get the feeling is happening.

    1. Anonymous10 months ago

      If the United States’ score is from the TOP 5 or ten percent, then their is definitely a problem.

  5. udit1 year ago

    no mention of china & India…………..

  6. Nathan Gustavson1 year ago

    As long as a business majors continue to make twice as much as engineers with half the education, why would Americans choose science? It doesn’t matter how cool the government tells kids it is if the economy doesn’t support it. Most of the engineers in my office are taking management courses to get out. We can’t keep the ones we have now..

    1. Adam1 year ago

      Engineers make far more than business majors by any measure. The jobs which possibly pay better than engineering jobs (although plenty of CS majors start at 140K all-in) are consulting and investment banking, and those are open to all majors, but engineers with a strong math background have a big leg-up.

    2. qwervqwe vqwevqwev10 months ago

      Business majors get paid well, but not as much as engineers.

      Also business is not always easy. The vast majority of 4 year colleges/universities require at least calculus I or statistics for any business major.

      1. Anonymous8 months ago

        So what you are saying is that all a business major has to have is differential calculus… Wow not really impressive…

  7. Mark2 years ago

    The problem for the US is we test ALL of your kids (the smart, the average, and the stupid) while the rest of the world only tests the smart kids. Of course the US results are Average, because of who we test. If we tested only to the smart and above average kids our test scores would be up with South Korea and the rest.
    This is the big secret nobody knows.

    It’s like having a dog barking contest. The world brings their best dogs….the US brings a mixture of dogs and cats. Then we get mad when the cats can’t bark. We spend millions and millions of dollars to teach cats to bark. No matter how much money you spend the cats will never, ever bark.

    1. Wcg1 year ago

      As far as I know Canada tests all students and I suspect Finland and other European countries do too. I don’t see a reason to compete with the rote learning systems of Asia but the West should aspire to the top ten.

      1. Kit1 year ago

        America doesn’t need to compete with the world, they just need to be able to afford to hire the rest of the world.

    2. Tony Zintnski1 year ago

      all countries are averaged, just guess Amercia’ are a little thicker than most western. Perhaps if the tests was on warmongering or fast food they would get a higher ranking.

    3. Anonymous8 months ago

      Canada tests all it’s students. How do you know this “big secret”?? How asinine. You would rather imagine that everyone else is cheating than think that your government is failing you and you are failing your children by demanding change.

  8. Bart Zehren2 years ago

    How would the country-level scores be different if the tested students were drawn as a random sample from ALL children of the school age range? That is, without regard to whether or not the selected child is, or ever has been, in school. Can you provide data in that way?

    Please let us know as soon as researchers have figured out how to administer that kind of measurement and what they’ve found.

    Also, of course, by showing proportions of students “performing” at different levels your report tells us nothing about the “quality” or “performance” of the schools in a country. What is learned is information about the mix of students living in countries as compared to one another at given points in time. These data do not report results of cohorts of students trakced longitudinally. Thus they are not about student progress at all, they’re about a static, cross-sectional comparison of fully separate respondent groups, i.e., of different subsets of students.

    If we’re going to focus on that perspective, how about showing the different amounts of variance of students’ scores across the student bodies tested in each country? Then, given my comment above about disregarding school attendance as a qualifier, one could mark, for analysis purposes, those who are attending school vs. those who are not, as well as note the relative proportion that school attendees represent in each country. I offer this in the spirit of “No Child Left Behind” so that ALL children are considered and represented in such global research efforts and reports – not just those currently enrolled in a school (not to mention, depending on the country, not just those enrolled in certain types or kinds of schools).

  9. Jon2 years ago

    This is more a ranking about standardization of education. Countries favoring rot learning without thinking are at the very top. For instance, France ranks pretty low in this ranking. Yet high schoolers who are good at physics/math go on to graduate with a Scientific Baccalaureate (end of high school) , with a level rarely attained by their Asian counterparts, and for that matter Americans with a 4 year college degree. This ranking therefore is deceiving and inaccurate. It takes population as a whole or a sample, not taking into account countries with an elitist education system like France, where one’s path is already decided in junior high school.

    In countries like Japan, tests are a lot more standardized and everyone is expected to learn the same things up to a certain age. However the scientific approach, that of learning how to do a proper mathematical demonstration would not score very high.

  10. alancook2 years ago

    National math test scores continue to be disappointing. This poor trend persists in spite of new texts, standardized tests with attached implied threats, or laptops in the class. At some point, maybe we should admit that math, as it is taught currently and in the recent past, seems irrelevant to a large percentage of grade school kids.

    Why blame a sixth grade student or teacher trapped by meaningless lessons? Teachers are frustrated. Students check out.

    The missing element is reality. Instead of insisting that students learn another sixteen formulae, we need to involve them in tangible life projects. And the task must be interesting.

    Project-oriented math engages kids. It is fun. They have a reason to learn the math they may have ignored in the standard lecture format of a class room.

    Alan Cook

    1. John2 years ago

      Your comment indicates a preference for STEM methodologies… do you believe we should go more in that direction?

  11. Ram2 years ago

    No assessment of China & India in this report.

  12. Richard Askey2 years ago

    PISA does not give a comprehensive summary of what goes on in mathematics education. To see an example see

  13. Meme Mine2 years ago

    *Quote one CO2 scientist that says the scientific method won’t “allow” them to say a “THREAT TO THE PLANET” is “proven”.
    Climate blame “belief” was to liberalism what the “right to life” issue is to the knuckle draggers of the Tea Party.