After more than a decade of heavy investment in closing achievement gaps and bringing all students to proficiency in reading and mathematics, the United States has fewer low-performing students on the Program for International Student Assessment—but only in science.
In math and reading, by contrast, there were no changes at all in the share of low-performing students on PISA between 2003 and 2012, according to a new analysis by the Organization for Economic Cooperation and Development. America was mostly flat during that period, remaining a little worse than the international average in the share of students who performed below minimum proficiency in all three subjects. Each of the three core subjects in PISA is administered together every three years to 15-year-olds in more than three dozen countries. The assessment tends to focus on critical thinking and ways students apply what they have learned.
Among U.S. students in that age group, 26 percent were low-performing in math, 17 percent in reading, and 18 percent in science. More than 1 in 10—some 95,000 students—scored low in all three subjects.
“These are big numbers,” Andreas Schleicher, OECD’s director for education and skills, said in a briefing with reporters. “You translate that into the future, these are people who will be underemployed, unemployed. ... This is a very significant liability for our society.”
Nine other countries did significantly reduce the number of students who were low-performing during the same time frame, including Brazil, Mexico, and Russia.
The OECD considers students “low performing” if they score below level 2—for example, less than 420 points on a 1,000-point scale in math. And American students didn’t always do well even on level 1 questions: Only 54 percent of U.S. students correctly answered a math question requiring a student to calculate an exchange between two currencies, which was set at a difficulty level well below level 2 and which 80 percent of students across the OECD answered correctly. In fact, out of 41 OECD countries, only Brazil had fewer students get the question right.
Science a ‘Puzzle’
In contrast to math and reading, the proportion of low-performing students in science decreased by 6 percentage points between 2003 and 2012. “I think the science result in the U.S. deserves some further analysis,” Schleicher said. “It’s a puzzle to us, a puzzle to me.”
OECD’s analysis, like many other studies, found that a student’s risk of being a low performer creeps up steadily from a host of disadvantages that vary in importance from country to country. For example, 80 percent of girls in poverty with other challenges performed below minimum proficiency in math.
Poverty was a factor everywhere, but its effect differed widely. In the United States, a student in poverty was seven times as likely to be a poor performer as a wealthy student, while in the OECD generally, poor students were four times as likely to be low performers.
Moreover, in the United States and 24 other countries with similar demographic and educational profiles, a student’s poverty increased the risk of other characteristics, such as being an immigrant or a girl, speaking a different language from the home country, or having had little or no preschool. By contrast, 21 countries including Brazil, Mexico, Tunisia, and Turkey, all showed that students in poverty with other risk factors had a lower likelihood of being low-performing, suggesting they had more supports for those students.
The OECD also found that while educational resources were needed to reduce a country’s pool of low-performing students, the amount of per-pupil spending in each country was not as closely linked with performance as with how equitably countries spent the money they had.
Students’ own dedication and confidence in their abilities played a big role, too, the OECD found.
For example, the OECD found students who completed six to seven hours of homework each week were 70 percent less likely to be low-performing in math, and those who participated in such extracurricular activities as art or music were even more likely to be proficient.
But the OECD also found that low-performing math students, wealthy or poor, were significantly more likely to believe that their efforts were meaningless and nothing could help them get better.
“Low-performers look alike in attitudes toward school, attendance, belonging, and math self-efficacy, regardless of whether they are from disadvantaged backgrounds,” Schleicher said. “Many students say, that’s all about talent, that’s all about things beyond my control.”
