Boys are outperforming girls significantly in three high-profile university admissions tests, raising questions about females' grade superiority at GCSE and A-level.
The findings come from trials of a US-style university entrance exam, from a well-known medical test for academic high-flyers, and from the history exam at Oxford University.
The most detailed results come from a UK pilot of the SAT test (so called because it is based on what was once the scholastic aptitude test), an assessment of reasoning in maths and English that has been used for decades to help US universities choose undergraduates.
The new test is being trialled by the National Foundation for Educational Research in the hope it will help universities identify talented pupils from disadvantaged areas.
But analysis, based on 8,000 pupils taking the tests in 2005, suggests there could be a strong gender effect if it is introduced.
Boys significantly outperformed girls as a whole, the academics found. This was due mainly to their performance in maths, in which they did better despite girls scoring higher in both GCSE and A-level maths. The males' average score was 523, against 480 for females. Yet 43 per cent of the girls had already gained A grades at A-level or AS, compared with 37 per cent of boys.
The picture was different in English, where girls and boys were roughly level in reading, but girls edged ahead in writing.
The academics also looked at the results based on what might be expected given each pupil's A-level grades. This found boys consistently performing above what would be predicted, while girls achieved below expectations.
Catherine Kirkup, lead author on the study, told a conference run by the Nuffield Review of 14-19 education that the findings were especially striking for maths.
"At every A-level grade, male students do better in the SAT than females. What this is down to is that the boys are much more prepared to have a go on multiple choice tests than girls," she said.
Evidence for this came from the fact that boys left fewer questions blank. Questionnaire evidence also suggested they tend to be more confident of their abilities.
Analysis of results from the biomedical admissions test, presented to the same seminar, also found boys scoring higher than girls. This test is set by five leading universities for entrance to medical school. It includes extensive multiple-choice elements.
But any suggestion that it might be biased against females and not recognise their potential is undermined because it has proved a good predictor of students' performance in degree courses.
However, Mark Shannon, presenting the Cambridge Assessment research, said the results could be explained by the fact that girls who apply tend to come with a wider spread of ability than boys.
Finally, John Watts of Oxford University told the seminar that, in Oxford's history exam, "there is a fairly significant gender bias, with more males scoring in the higher bands".
The findings will fuel the debate about the gender gap in education. Girls are ahead of boys in most GCSE and A-level subjects, but international evidence presents a more mixed picture. Two weeks ago, The TES reported that Professor Alison Wolf, of King's College London, had argued that GCSEs and A-levels may favour girls over boys.
Earlier this month, it was revealed that the SAT appears to be failing in one of its central purposes of identifying the talents of poorer pupils. Children from disadvantaged backgrounds tend to do worse on the test than might be expected given their A-level results.
WHY SHOULD DILIGENCE TRIUMPH OVER QUICK THINKING?
The latest findings will no doubt prove appealing to male conspiracy theorists. They will argue that the latest test data show what has long been suspected - that mainstream exams, around which England's education system revolves, systematically favour girls.
With their emphasis on working consistently towards carefully defined goals and hard work over a long course, exams were bound to reward the more conscientious, who tend to be female.
But change the test - to focus on quick thinking and risk-taking - and you get a different result.
It is hard to quibble with most of these conclusions given the detailed findings of the SAT study.
In questionnaire surveys, sent to those who took the test after they had started their university courses, more males than females reported confidence in their abilities.
However, more females than males said they worked hard at school.
To state categorically that boys are being discriminated against in exams, though, is a tough call.
It seems likely that the designers of GCSEs and A-levels are simply choosing to accentuate qualities, such as diligence, that they think are important in all students.
In other words, boys are, on average, getting the results they deserve if they choose to work less hard than girls.
That said, the university entrance test scores suggest that other qualities may be of use in later life, particularly at university.
The latest data deserve further study by those responsible for setting exam standards. Perhaps Ofqual, the new testing regulator, would like to take a look.
EXAM DISCRIMINATION AGAINST GIRLS IS NOTHING NEW
Hurrah! We've found a test where boys do better than girls! All is right in the world once again.
After all, if boys don't do well, there's something wrong with the system.
In fact, girls have been outperforming boys in public exams since . well . the introduction of public exams. And we are as keen to devalue their achievements today as we were in the 19th century. ("So girls do well in literature?" the frock-coated educationists fulminated. "In that case, literature isn't a proper subject.")
But let's look at what the US-style test actually is: pupils are presented with a question designed to test their reasoning, followed by a selection of four answers. This is great if you want to test someone's ability to memorise facts, to take a wild guess, or to think within confines carefully delineated by a - probably male - examiner. In other words, it's great if you're a boy.
Of course, it would be reductive to say that fact-memorisation or reckless risk-taking are exclusively male preserves, or that intuitive interpretation is exclusively female. But we are culturally conditioned to value certain qualities in men and others in women.
We are also culturally conditioned to devalue any skill that is labelled "feminine". So when girls do well in exams, the response is to see what we can do to make them more boy-friendly. Key stage 2 girls have been forced to write about lizards, insects or fighter pilots for years, in an effort to boost boys' writing scores.
The new tests will undoubtedly be seen as proof that it's the education system that's been at fault all along.
And now we can overlook all those pesky questions about unequal pay for women or concrete ceilings in the workplace, safe in the knowledge that boys are cleverer, and therefore more deserving.