Boys and girls are similar yet profoundly different. There is no news in that, of course. Even in the 1920s, the Board of Education concluded that the two sexes were of the same intellectual ability and deserved the same curriculum, but that girls were more conscientious. Boys were described as prone to "healthy idleness".
None of this was held to matter, despite girls' wholesale failure to achieve their potential. Now, though, the picture has been reversed and girls are outscoring boys in almost every subject at every stage in the education system (except for first-class degrees at Oxford and Cambridge).
Suddenly, with boys no longer in the driving seat, the gender gap is hot news.
In some ways this is puzzling. Overall, boys' results have been improving, even if girls have been improving faster. At the end of the day, moreover, men continue to hold down the top jobs. But there is a wider concern. Boys'
reluctance or inability to work as hard or effectively as their sisters has been associated in the public mind with a more obviously troubling phenomenon: the disaffection of low-achieving boys, their increasingly disruptive behaviour, their exclusion from school and the growing numbers who end up in the criminal justice system.
This is why the issue has given such a high priority by the Department for Education and Skills, and why a thousand researchers have been unleashed to find ways of bringing boys back on board.
What are the inbuilt differences?
Research suggests girls have superiority in some verbal abilities, notably among pre-school children and adults. Boys, on the other hand, appear to have higher numerical abilities. These emerge in adolescence, then increase, particularly on tests of mathematical problem-solving. According to global research on 15-year-olds carried out in 2001 by the Organisation for Economic Co-operation and Development, girls are better readers, but in maths boys are superior in most countries.
Gender differences and young children
Differences are apparent even before formal schooling, but they are closely tied up with social expectations. Researchers who looked at pre-school children's choice of activity found girls drawing, reading or talking to adults, while boys were more likely to play with construction toys. There were also differences in attitude. The girls were more ready than the boys to ask for help with a jigsaw puzzle, although the boys seemed more confident. Yet there was no difference in ability to do the puzzle.
These early roles, quoted in research by Professor Jannette Elwood from Queens University, Belfast, have potentially far-reaching consequences.
Boys who pretend to be firemen and girls who dress up as "grand ladies" will have their own imaginative takes on the world.
A 1995 study of primary children's attitudes found different approaches even to a simple game of Lego. Girls made houses, boys built gun batteries; girls used collections of bricks as part of a social game, boys made complicated machines.
What part do expectations play?
A huge one. Little boys are deemed more able than little girls, despite evidence to the contrary. When adults were shown videos of children playing with jigsaw puzzles in the nursery mentioned above, they concluded that the boys were more confident, and so better. When it turned out that the girls had done just as well, this was put down to luck.
Such attitudes seem to be shared by the children. Throughout life, males are shown to be over-confident, believing they can do well with little effort. Girls and women, in contrast, consistently underestimate their abilities.
The nursery experiment also showed that teachers adopt gender-based strategies. They said they had to "capture" boys' attention by pandering to an interest in, say, technical toys, whereas girls were seen as more in tune with typical nursery activities. Girls' lack of interest in mechanical toys was not seen as a problem.
A 1999 study by the National Foundation for Educational Research (NFER) concluded that teacher attitudes probably had more effect than any inherent difference in linguistic abilities. Boys, then, are entering primary school disinclined to settle down; lacking interest in stories that deal with people rather than objects; and ill-practised at holding a pencil because they have spent all day racing round. As Professor Elwood suggests, they are likely to be at a disadvantage when faced with programmes for reading and writing.
How do the differences develop?
Even ignoring slight differences in linguistic and numerical abilities, boys and girls seem to have fundamentally different approaches to answering questions. They perceive different problems and find different solutions.
Girls, more than boys, value the circumstances in which tasks are set, and tailor their answers accordingly. Boys consider questions in isolation, judging the context irrelevant. Professor Caroline Gipps at Kingston University identifies boys' refusal to adopt collaborative learning strategies as a factor in their falling behind, particularly before 16.
Gender differences at secondary school
When pupils choose their timetable in the sixth form, clear preferences emerge. Boys predominate in some A-level subjects. In physics, for example, there were 23,595 male entries last year and 6,988 female; business studies and economics are also heavily male-dominated. But girls (55,451) hugely outnumber boys (23,295) in English. In French, too, women outnumber men two to one.
When they arrive at secondary school, boys' and girls' average levels of performance are fairly close, although girls do better in tests of written English. At age 11, 75 per cent of girls reach level 4 in English key stage 2 tests, against 70 per cent of boys. In maths, the performance is almost even, with 73 per cent of boys and 72 per cent of girls reaching level 4.
This slight gap in performance grows as secondary education continues.
In last summer's GCSEs, 62.4 per cent of girls gained five or more top grades, compared with 53.6 per cent of boys. At AS-level, 58.9 per cent of girls achieved grade C or above, compared with 51.4 per cent of boys. At A-level, 70.7 per cent of girls got A-C grades, compared with 63.8 per cent of boys.
At A-level, girls now account for more entries than boys and outperform them in almost every subject, including the sciences. There are anomalies at A-grade, for example, where boys outscore girls in English. This is because the boys are a comparatively select group. The same is also true of girls in physics, who, again, score more A grades.
Whether an overall gender gap has widened significantly since the early 1990s remains a matter of dispute. Research led by Professor Stephen Gorard at the University of Cardiff found that differences related to certain subjects - particularly English - among certain groups of pupils.
English, maths, and science
Although held to be a "boys" subject, the gender gap in science has decreased over time, suggesting that social pressures played a part in the past. Evidence from the Government's assessment of performance unit (APU) suggests girls outpoint boys on practical tests and on interpreting observations. But boys are better at applying physics concepts, a comparative ability that increases with age.
In English, APUresearch suggests girls' command of written language is the most important factor in their higher test scores - a phenomenon visible in key stage 2 Sats. It is also clear that boys and girls want different forms of knowledge from reading material.
In maths, surveys by the International Association for the Evaluation of Educational Achievement (IEA) reveal little difference in overall performance between 13-year-old boys and girls in maths, although girls tend to do better in computation, whole number and algebra, while boys score better in geometry and measurement. At 15, the gap grows, with boys ahead. Shifts in patterns of achievement, though, suggest social factors may be at work.
Differences at university
Here again, women are doing better. In 2000, 23 per cent of 18-year-old girls, but only 18.4 per cent of boys, went to university. More than half, 55 per cent, of graduations in 2002 were female, and they got more first and upper-second class degrees. Only at Oxford and Cambridge does the picture change, with men gaining the majority of firsts.
What's the effect of exam style?
Although it is often said that coursework favours girls, it is not clear that, overall, this is enough to give them a decisive advantage at GCSE.
Some research suggests it may do little more than counteract the disadvantage of the traditional "confrontational" exam setting, which is held to favour boys.
Most people agree that boys are happier with clear-cut questions and girls with more discursive approaches. Boys appear to be significantly happier with multiple-choice approaches, much favoured at the old O-level, with the inherent encouragement to take risks.
What's all the fuss about "tiering"?
Tiered maths papers give students and teachers three options at GCSE:an easier exam in which students can achieve no higher than a D, a middle paper for which B is the top grade, and a harder paper in which it is possible to achieve an A*. This allows flexibility. But there is some concern that girls' lack of confidence and teachers' low expectations have led to a disproportionate number of girls being entered for the two lower tiers of the exam. This is particularly significant as most universities require at least a C grade in the subject.
The Qualifications and Curriculum Authority is piloting a scheme that allows students to be entered for the top and middle papers, increasing the chance of a "pass".
"It's at the top level that there's a big difference," says Margaret Brown, professor of maths and education at Kings College, London. "Girls in the top set told us the speed of coverage in the syllabus was too quick. Girls were asking to go down a level and feeling depressed about their progress.
The boys reported this too, but worried less about it and saw it as a game."
What is the effect of teaching strategy?
Several strategies have been used but, as NFER points out, they have often been introduced without any evidence that they work. At primary and lower secondary levels, schools have been able to select reading material to appeal to boys - a mixture of child-friendly horror and machines - with some success. Some, including Dr Debbie Epstein, co-editor of Failing Boys?Issues in gender and achievement, have questioned the wisdom of this approach, arguing boys are already under enough pressure to be rough and tough. Other strategies include boy-girl seating to prevent pockets of misbehaviour, and single-sex classes. Boys, it is often said, hog the attention in mixed-gender classes and can lose focus. Girls, meanwhile, can benefit from single-sex teaching because they are less prone to gender stereotyping.
Firm conclusions are hard to find. Government-backed research into helping boys improve at secondary school has found that organisational changes are not enough. Instead, straightforward good teaching, including the development of effective relationships between staff and students, appears to be the basis for improvement.
Researchers from Homerton College, Cambridge, looked at tactics that appeared to help boys without disadvantaging girls. Some schools had altered their teaching styles, breaking up lessons into smaller chunks and incorporating five-minute "breathers" for boys. But for every school where innovative strategies helped, there were others where they made no difference, or made the situation worse.
Researchers from Edinburgh University have suggested that boy-girl seating does reduce disruption. But girls dislike the arrangement and neither sex picks up anything from the learning styles of their opposites.
What about single-sex schools?
Since the introduction of league tables, girls' schools have used exam results to justify single-sex education. But in 1999, researchers at the Institute of Education found that social class, ability and the school's tradition had a much greater impact on girls' results.
This finding matches those of other researchers in Britain and abroad.
There were other effects, too. Girls' confidence was improved at a single-sex school; yet there was also an unhealthy "spiteful" aspect to the competitive atmosphere.
Will recruiting more male primary teachers improve boys' results?
Maybe not. Research published in 2002 by Hertfordshire University found no link between the number of male teachers in a school and the performance of its pupils in key stage 2 tests; the proportion of pupils with special needs was more significant than the teacher's sex. Researchers at the University of the West of England have reached similar conclusions, pointing out that the imbalance between male and female teachers in primary schools is of long standing. Men represented only 25 per cent of elementary school teachers in 1914 and the same proportion of primary teachers in 1967. The proportion has fallen since, but only slightly, to 22 per cent.
What else could make a difference?
There are some points of consensus. The need to combat the "laddish", anti-learning culture among some boys is one. The need to avoid stereotypes is another; it is no more acceptable for girls to lack confidence than for boys to misbehave. HMI has identified common characteristics in schools where boys succeed, including: a non-macho culture, a strong sense of community, an ethos that values achievement, where teachers provide prompt, detailed responses to pupils' work, and where the teaching is enthusiastic and delivered with humour.
Boys could probably help themselves, too. Professor Gipps told last autumn's meeting of the Headmasters' and Headmistresses' Conference that they probably needed to develop some of the collaborative questioning learning strategies used by girls.
Does it matter?
Factors such as poverty and race continue to be far more influential on a child's eventual fate. Middle-class boys may do worse than middle-class girls, but they still do better than working-class girls.
And for many pupils, exams are the least of their worries. "For both groups, we have to acknowledge that becoming an adult is the key concern in the secondary years," says Professor Gipps, "and this can often take over from, or even get in the way of, studying and learning."