Michael Gove’s 2014 A-level reforms offered a sweeping and apparently profound solution to systematic problems. Coursework, except where absolutely vital, was consigned to history; AS levels were scrapped in favour of a single set of tests at the end of two years. The then Education Secretary said the changes would introduce greater rigour, eliminate cheating on coursework, and revive the “art of deep thought”.
In the heated debate around these reforms, one thing that hasn’t received much attention is whether the new format affects all students equally. In 2017 – the first year where students had taken the new A levels – boys achieved more A* to A grades than their female classmates for the first time since 2000. On A* to C results, where girls have typically benefited from a sizeable gap, their lead shrank for the first time since 2010.
In reformed subjects like history, boys increased their share of top grades; the share achieved by girls, again, fell. If this trend continues, then as reforms are applied to the remaining subjects, the female advantage in compulsory education could be rapidly reversed. Though not immediately concerning, it should make us think about how small changes in the way we assess and teach our children can have huge implications for who succeeds and fails.
The widely perceived view is that girls are better at coursework, while boys are better at exams. It is certainly true that girls do more homework: an hour more a week, according to research. That suggests a level of preparedness which perhaps maps more neatly onto ongoing assessment such as coursework than the cramming for a single big exam. Mary Bousted, General Secretary of the ATL, has said that girls prefer more continuous, “measured” assessment, while boys prefer to have it all in one go.
But are these gender stereotypes any more than an urban myth?
Yes, according to research led by economist Ghazala Azmat. According to a report for the LSE’s Centre for Economic Performance, imbalances in performance are related to the pressure of different forms of examination. Azmat assessed students in relation to a series of high, medium and low-pressure exams over a six-year period and found that girls performed significantly better in low-pressure classroom tests, while boys edged ahead on more intensive national exams.
The difference is even starker when you consider that girls did almost twice as well in low-stakes tests as they did in high-stakes ones. As Azmat herself notes, this may mean that high-pressure exams simply don’t do female students justice – despite working harder, and more consistently, for the rest of the academic year.
The effects of this difference in performance are potentially far-reaching. Intuitively, people are much more likely to carry on doing courses that they score well in. It happens that STEM subjects, which have historically been taken up less by female students, have tended to rely almost exclusively on final examinations. In contrast, subjects such as History and English Literature, more popular with girls, are much more coursework-heavy. Could worse results than expected in STEM exams be having an effect on uptake later down the line?
Perhaps; different choice of subjects might also be partly explained by higher reading abilities among female students. But the thinking corroborates with 2015 research by Pisa, which suggested a strong link between anxiety, which was higher for girls, and lower performance in STEM assessments. Female students routinely underperformed in exams, despite doing better in classwork. If true, it not only busts the myth that boys are ‘naturally’ better at STEM, but it might also go some way to explaining why consistent attempts to get more girls into these subjects have failed to hit the mark.
Around half of this year’s A-Levels were set in the new, exam-heavy format, including many of the most academic ones such as History, English Literature, and the sciences. From autumn almost all subjects taught will be in the new format. Coursework is out; exams, and cramming, are definitely in.
This is certainly not all bad; cheating on coursework has become worryingly routine. Methods have included letting pupils work for longer than allowed, hinting at what material will be assessed, marking essays up, and providing ‘template’ essays that pupils can essentially copy. To make matters worse, cheating in this way has also been shown to widen the achievement gap for the poorest pupils. In 2016, 388 penalties were issued for cheating on GCSEs and A levels. This is huge, and yet still likely to be a significant understatement of the true extent of the problem.
Nonetheless, unintended consequences are becoming apparent. The move away from modular and regular assessment, evidently, will reward those who are good at exams over those who may be better at coursework. It also seems reasonable to imagine that a new, higher pressure exam period will increase the often significant anxiety faced by pupils.
With the scrapping of AS levels too, this means that for most subjects the final exams will be the only assessments for the course, which makes them the highest of stakes. The government has consistently said it wants more girls to get into STEM. Presumably, it also does not want to unfairly advantage one gender over another.
It is not yet obvious what effect the full reforms will have on the gender divide, but things will become a little clearer after results are released this August. Gove brought us a comprehensive solution to a complex problem, but we should think carefully about the unintended consequences of his reforms.