Women go into science careers more often in countries without gender equality

A large number of social factors have discouraged women from pursuing careers in science and technology. But in a number of countries, an increasingly egalitarian view of gender differences has been associated with rising math and science scores for girls. However, that change hasn't been followed by increased participation in science and tech careers; in fact, the frequency of women pursuing degrees in these areas is often higher in societies that are far from egalitarian.

Two researchers, Gijsbert Stoet of the UK and David Geary in the US, decided to explore this paradoxical trend. Their analysis suggests that the situation may be the product of a complex mixture of relative talents, general confidence, and social factors. The results drive home that, if we want to attract and retain some of the best talent in the sciences, it's going to take more than simply ensuring they have equal access to advanced degrees.

Stoet and Geary's research relies on a lot of publicly available information. One of the keys to this work is the Programme for International Student Assessment, or PISA, which gives standardized tests to students around the world. The most recent iteration of these tests was given to about half a million students in a total of 71 countries and, so, provides a trans-national measure of students' skills in math, science, and reading comprehension. Critically, when it came to science, the PISA survey also asked about students' interest in and enjoyment of science, as well as if they felt confident they could do some basic scientific analysis without supervision.

To understand the context of a country's test scores, the two researchers also obtained information on gender gaps in different countries from the World Economic Forum, educational data from UNESCO, and the overall life satisfaction in the country from the United Nations Development Programme.

Overall, the data showed that the gender gap in science is pretty small among the children tested. While boys outperformed girls in science in 22 countries, girls came out on top in 19. There was no clear relationship between these results and a country's gender equality.

But things got complicated when Stoet and Geary looked at each student's relative strengths. This involved averaging their scores for all three subject areas, then comparing the score in each subject to the average to identify the student's strongest subject. Thus, someone could be below average overall but still have math be their strongest subject or could be well above average in reading comprehension yet be relatively better with science. Among boys, it is far more common to have science be their strongest subject—this was true in every country but two (Lebanon and Romania are the outliers), which means it was also true in most of those 19 countries where girls on average outperformed boys on the science test.

In math, this analysis favored boys in every single country; in reading, the converse is true, with girls coming out on top in every single country. As a result, even in situations where girls outperformed boys in science, they typically outperformed them in reading by an even larger margin.

It might seem that this alone could account for the relatively low percentage of women pursuing a career in science and technology—but it doesn't. In every single country, the percentage of women getting degrees in relevant fields was lower than the percentage who had science or math as their strongest subject. So, even if people were being drawn to get degrees in the areas they were best at, it wouldn't explain the gender gap among the graduates.

So what could explain this? There are hints it comes down to social factors. In 76 percent of the countries, boys expressed a bigger interest in science, and this effect was especially pronounced in countries with high gender equality. In addition, they were more likely to express confidence that they understood the scientific process well enough to do science without external guidance. Again, this was true even in countries with high gender equality.

This doesn't mean they could do research without assistance, or even really felt they could—only that they said they could when asked. In fact, the researchers checked and found that ratings of self-sufficiency were only weakly correlated with good test scores. When done on an individual basis, boys' self-sufficiency ratings were higher than their test performance in about half of the countries; for girls, this was only true in seven percent of the countries, which suggests that boys' confidence isn't entirely justified.

What this doesn't explain, however, is why the gap tends to be larger in countries that have high levels of gender equality. But Stoet and Geary have an idea here: they note that these countries tend to have very high levels of economic security, with a large social safety net. The countries where the science-degree gender gap is smaller tend to be less socially secure. The researchers suggest that the economic security provided by fields like engineering may have a stronger draw in these countries, pulling more women into the field.

They attempt to use a statistical pathway analysis to see if the data is consistent with this being the case, but the results are inconclusive. It may be right, but there would be at least one other strong factor that they have not identified involved.

Still, the researchers have found two factors that might explain some of the gender gap in people getting science degrees: boys are more confident in their science abilities (even when that's not justified); and even in cases where girls have better performance in science, their performance in reading is typically even larger. But it's clear that there's more going on. "If absolute performance, interest, joy, and self-efficacy alone were the basis of choosing a STEM career," Stoet and Geary write, "we would expect to see more women entering STEM career paths than do so," which means we still have some work to do if we want to draw some of the most talented people into the scientific endeavor.

Psychological Science, 2017. DOI: 10.1177/0956797617741719 (About DOIs).