We all know about the gender imbalance in the uptake of Stem (science, technology, engineering and mathematics) subjects at school, university and beyond. A couple of months ago, UK business secretary Vince Cable highlighted the fact that this country has the lowest percentage of female engineers in Europe.
But to be clear, this is a problem that is near-universal.
Girls do at least as well as boys in these subjects up to the point where they get to choose routes out of them. So it's not about girls underachieving once they have opted to specialise, it's about why they don't opt for specialist Stem subjects in the first place. It's about "sticky floors" not "glass ceilings".
Gender stereotyping is an issue. School students are making subject choices at a time when they are developing their identity and exploring their self-image. Crucial to this is how they see themselves in relation to the subjects they are studying.
In addressing gender imbalances in subject take-up, the imperative is not to water down science or maths, nor to attempt to strong-arm unwilling students into Stem subjects. Instead it is to identify and remove any obstacles to unconstrained choice. Then, if a student still prefers Middlemarch to molecules, at least it won't be because of a misplaced sense that one is more gender-appropriate than the other.
So where are these constraints? They lie in the classroom (around what is taught, where it is taught and how it is taught); but they also lie outside the classroom, within the overall school environment.
What is taught
Many experts argue that girls prefer to study topics in context. They appreciate understanding their relevance and connections to other topics and fields, whereas boys are often happy studying abstract concepts in isolation. Too often this is taken to mean that for girls to enjoy science it has to be "watered down" by including a social element. But making links between topics and subjects can be a higher-order thinking skill.
Another aspect is that of role models. The historical contributions of women scientists and mathematicians tend to be under-exploited - think of Rosalind Franklin.
Where it is taught
The fact that cross-curricular initiatives tend to be in the humanities is a shame for science. Among scientists in academically inclined schools, too, there is often opposition to moves to make key stage 3 science (for 11- to 14-year-olds) "general", rather than introducing the three sciences as separate strands. But increasing numbers of specialists advocate teaching students how to be scientists before introducing them to the different ways in which biologists, chemists and physicists explore the world. This study of science as science can also lead to more innovative ways of opening up the subject.
How it is taught
Experts suggest that not only do girls typically engage better with science when it is taught in context but they also relate well to collaborative, project-based and enquiry-based approaches to teaching and learning. The crucial issue is that of classroom dynamics. Considerable research and anecdotal evidence suggests that in co-educational classrooms, gender stereotyping operates insidiously to disenfranchise girls. One study found that in physics classes, girls made faster progress when paired with other girls than they did in mixed groups. Conscious efforts need to be made to avoid gender-biased expectations, to encourage girls to take an active role in practicals, and to get them to ask questions and volunteer answers.
Overall school environment
The steps teachers take in classrooms will be most effective when the school culture as a whole supports initiatives to engage girls with Stem subjects. Raising the profile of science itself - through the identification of appropriate role models and communicating the career options open to scientists of both sexes - helps to reinforce a refusal to accept that some subjects are more feminine than others.
Making Stem subjects more attractive to girls is a complex task. Single-sex settings can be helpful, whether at a whole-school level or in particular subjects and key stages, but this needs to be backed up by an inclusive curriculum, an effective and engaging pedagogy and high teacher expectations. Inspirational teachers are crucial, but the whole school needs to buy in to what is being attempted.
Most importantly, a rejection of preconceptions is required. The idea that science is peculiarly difficult, abstract and content-driven plays to the image of it as exclusionist and "masculine". Making science more engaging, inclusive, contextual and contingent does not mean making it easier. If anything it makes it more challenging, more fulfilling and more gender-neutral.
If more girls are given the impetus to continue their study of Stem subjects, they won't be the only ones to gain: science itself will benefit, as will the whole of society.
Dr Kevin Stannard is director of innovation and learning at the Girls' Day School Trust.