How Not to Close a Mathematics Sex Gap

We’ve been pondering this one for a year, not quite sure how to handle it. The off-blog crusading and being flat out with Jekylling also provided the excuses to keep putting it off. But, it’s gotta be done, so let’s quickly do it.

Last year, a very big study appeared in Nature:

Rapid emergence of a maths gender gap in first grade

This study reported on the testing of roughly a zillion French first and second grade students. From the abstract,

Boys and girls exhibited very similar maths scores upon school entry, but a gender gap in favour of boys became highly significant after 4 months of schooling and reached an effect size of about 0.20 after 1 year. These findings were repeated each year and varied only slightly across family, class or school type and socio-economic level. Although schooling correlated with age, exploiting the near-orthogonal variations indicated that the gender gap increased with schooling rather than with age.

Why? One could hazard a guess or two, of course, which would be, um, hazardous. In a discussion of the potential causes, the authors are careful to note that their study is purely statistical, and the “why” is not theirs to declare:

Given the observational nature of our data, any attempt to infer the mechanisms producing the maths gender gap must remain hypothetical, …

The authors, however, don’t really believe this:

… but some explanations are, nevertheless, more consistent with our findings than others.

The authors repeatedly ignore their own hazard warning, beginning with the first sentences of the Abstract:

Preventing gender disparities in mathematics is a worldwide preoccupation. In infancy and early childhood, boys and girls exhibit similar core knowledge of number and space. Gender disparities in maths are, therefore, thought to primarily reflect an internalization of the sociocultural stereotype that ‘girls are bad at maths’.

Yeah, so this is all due to stereotyping in the classroom. After all, it’s not like much else changes in the first years of a kid’s schooling.

For such reasons [the inconsistency of earlier signs of sex differences], young children’s mathematical attitudes, perceptions, interests and competence are thought to constitute universally shared ‘neurocognitive start-up tools’ that are later shaped by a sociocultural belief that girls exhibit lesser proficiency in mathematics relative to boys.

Yeah, maybe. Or maybe there is more to innate mathematical ability and/or desire than neurocognitive start-up tools.

However, where, when and how widely this stereotype becomes entrenched remains uncertain.

So, one might ask “where” or “when” or “how widely”, but not “whether”.

The lack of average gender differences at the beginning of grade 1 also gives little credibility to the idea that the emerging gender gap is related to fundamental gender differences in aptitudes.

Um, why, exactly?

Our data do not, however, exclude more complex interactions between nature and nurture, for instance a latent advantage of boys in maths learning that comes into play only at the onset of formal education or of gender differences in competitiveness and test-induced anxiety that are exacerbated by school entry.

So, boys don’t have a greater “aptitude” but may just be more into “competitiveness”. The authors could have employed a less win-lose word, such as “challenge”, but it’s authors who are really about the competition. And, of course, “maths anxiety” has gotta be part of any explanation:

Girls may exhibit greater maths anxiety, possibly reinforced when facing competitive tests, a behaviour that may explain why, among all maths and language exercises, the male advantage is more pronounced in more challenging, new or complex tests tapping executive functions. This explanation is congruent with our finding that greater test difficulty enhances the gender gap.

Girls also suffer more from maths anxiety than boys, especially in scenarios involving competitive or time-limited maths tests, an effect discernible as early as second grade worldwide.

In summary,

For such reasons, young children’s mathematical attitudes, perceptions, interests and competence are thought to constitute universally shared ‘neurocognitive start-up tools’ that are later shaped by a sociocultural belief that girls exhibit lesser proficiency in mathematics relative to boys.

The reporting of the study was largely done with this framing. Harvard, the academic home of Elizabeth Spelke, one the authors, contributed a puff piece:

Large-scale study adds to mounting case against notion that boys are born better at math

That is, a sex disparity that emerges at pretty much the first instance that the kids are being taught and tested on mathematics is “mounting” evidence that boys may not be better at maths, or may not simply, innately, care more to do it.

The overall results, the most conclusive to date, suggest it’s time to shelve explanations based on biology or bias. Instead, it appears there’s something about early math instruction that produces gender disparities.

“We still don’t know what that is exactly,” said Spelke, …

Perhaps Spielke, who came to her “strong position” some “twenty years” before the study, doesn’t care that much to know. Nonetheless, she and her coauthors know what to do:

Which factors should be targeted? Although our evidence is only correlational, we found that class-level variables, such as class size, gender ratio, heterogeneity in maths level or gender of the student at the top of class, have only a small modulating influence. Given that our data indicate that the maths gender gap starts with entry into the school system, improving teacher training should undoubtedly be one of the most powerful levers. Encouraging teachers’ gender-fair ratings and active cooperative practices such as questioning girls and boys equally often during maths and science instruction and focusing equally on the talents and efforts of children of both genders, are efficient practices that should become part of teachers’ basic training. Boosting teacher training in maths to increase their confidence and interest in this topic could also be effective, especially in countries such as France where most primary school teachers are female27. Interventions can also convince boys and girls that maths is worth the effort by exposing children to both male and female role models with whom they can identify; providing girls with ways to cope with competitive stress and maths anxiety; emphasizing an incremental view of intelligence in efficient learning4; and implementing self-affirmation tasks to protect girls from stereotype threats.

Whatever. None of this is too awful but none of it has much to do with anything. One can affirm and self-affirm until the cows come come, but it won’t make much difference on learning or excelling in mathematics.

God knows if this apparent sex gap is largely due to innate differences. Then, if the sex gap is a cause for concern – and that should not be taken for granted – God knows how one might narrow the gap. But it is very clear that the way to not narrow the gap is to game the framing of raw data, pre-perverting any discussion and analysis of any plausible policy.