By Julie Rehmeyer

Web edition: June 5, 2008

Three years ago, Lawrence Summers speculated publicly that biological differences between the sexes were the primary reason for the low numbers of women at the highest levels in science and engineering. The ensuing fury led to his resignation as president of Harvard – and to renewed interest in the science of gender differences. Girls have, on average, consistently scored lower than boys on many test of mathematics, for example the math SATs. But the big question that Summers’ remarks highlighted is whether those differences are cultural or biological.

New research shows that in countries with greater gender equality, the gender gap in mathematics closes. The research doesn’t, however, argue that biological differences play no role in academic performance. In every single country, girls score substantially higher in reading than boys do. The girls also tended to do better in arithmetic than geometry in all countries.

Paola Sapienza of Northwestern University in Evanston, Ill., and her colleagues analyzed the results of a newly developed, challenging mathematics and reading test for 15-year-olds, administered in 40 countries in 2003 and designed to be free of cultural biases. On average, girls scored 2 percent lower on the math portion and 6.6 percent higher on the reading portion than boys did. The U.S. scores were close to this average, but those percentages varied widely from country to country: In Iceland, for example, girls scored more than 3 percent higher than boys on the math section, whereas in Brazil, they scored more than 4 percent lower.

“These wide differences call for some kind of cultural explanation,” Sapienza says. So she and her colleagues correlated the test scores with several different scores of gender equality in each country. The World Economic Forum’s Gender Gap Index scores countries on economic and political opportunities, education and well-being for women. The World Values Survey asks people around the world questions like “When jobs are scarce, should men have more right to a job than women?” The researchers also used ratings of women’s economic and political activities. Countries that scored well on one of these ratings tended to score well on the others also.

In countries with high levels of gender equality, the math gap disappeared or sometimes even reversed, the researchers found. Countries with low levels of gender equality had larger gender gaps in mathematics. These results held up even when the researchers controlled for the economic development levels of the countries. “The gender gap in math is not there in every society,” Sapienza says, “so we should not think that it’s a permanent feature.”

The Macao region of China, Brazil and Turkey had low scores in gender equality and correspondingly low girls’ math scores. The Scandinavian countries and the United Kingdom were especially high in both. More surprisingly, Thailand and Indonesia, with middling scores in gender equality, had high girls’ math scores.

The girls’ success in gender equal societies didn’t come at the expense of boys: Everyone’s scores were higher, in both math and in reading. The girls’ gains, though, were greater in both areas, reducing the math gap and increasing the reading gap. “It would be a mistake to interpret this as bad news for boys,” Sapienza says. “We’re simply saying the girls are catching up.” She also notes that the boys in the study report spending far less time on homework than girls do, perhaps helping to explain their relatively weak performance in reading.

The researchers also studied the percentage of students of each sex among the top scorers on the test. In the gender-equal countries, girls made up half or more of those who scored in the top one percent. The sex ratio of top performers is especially important because these students are the ones most likely to excel in careers in science and engineering. Summers had suggested that because of biological differences, it is nearly inevitable that a much higher percentage of these top performers will be boys.

The researchers also noted that even in gender-equal countries, girls’ scores in arithmetic tended to be higher than in geometry. Previous research has suggested that boys tend to be stronger at spatial visualization than girls. Sapienza and her colleagues published their results May 30 in Science.

Sapienza is now interested to understand why girls perform better in gender-equal societies. Possible explanations include better role models, different incentives or different curricula. She also points out that even in gender-equal societies, girls may or may not choose to go into science or engineering. To enter those fields, she says, “You need to know that you can do it but also that you like it. Our data doesn’t say anything about that.”

Diane Halpern, a psychologist at Claremont McKenna College in California, cautions that “anyone who is looking for a single, simple answer about sex differences is going to be very frustrated.” Different tests can bring different results. For example, girls tend to do better on tests that are more closely tied to the curriculum, so there’s usually no gender gap on math scores on such tests to begin with. She also cautions that gender equality itself is a slippery notion.

But Elizabeth Spelke of Harvard University says that the study brings out the interrelationship between cultural impacts and biological differences. “The research also shows nicely how the genders can differ in their profiles of cognitive strengths and weaknesses, on average, without one gender being ‘better’ than the other,” she says.