Natalie Parletta
Natalie Parletta is a freelance science writer based in Adelaide and an adjunct senior research fellow with the University of South Australia.
An analysis of global data should shatter gender stereotypes that suggest females have an inferior grasp of STEM (science, technology, engineering and maths) topics.
The study, published in the journal Nature Communications, shows that females are better than males at sustaining their performance on longer tests involving maths and science, hence reducing the gender gap.
This unexpected discovery was made by researchers Pau Balart, from the University of Illes Balears in Spain, and Matthijs Oosterveen from the Erasmus University Rotterdam in the Netherlands.
They were “fascinated” by research that found student performance drops during test-taking, says Balart, and thought gender differences could yield interesting insights, especially if combined with already known differences in cognitive domains.
This notion offered an opportunity to challenge beliefs like that of Oxford University dons.
After giving females 15 minutes longer on maths and computer science exams, the dons declared a commonly held view to The Telegraph that “female candidates might be more likely to be adversely affected by time pressure”.
Balart and Oosterveen proposed another explanation: “female students might make better use of the extra time on the test because of their ability to sustain performance”.
Putting this to the test, they analysed data collected every three years from 2006 to 2015 in 74 countries for the Programme for International Student Assessment (PISA), a standardised assessment of 15-year-old students’ performance in reading, maths and science.
Results on early test items aligned with previous findings that girls do better with reading and boys with maths and science.
But when comparing each gender’s scores at different stages of test taking, girls indeed showed less decline in accuracy over time on all the tests.
Although boys had an initial advantage in maths and science, the authors write, “there was not a single country in which they were significantly better able to sustain their performance during the test.
“This finding suggests that longer cognitive tests exacerbate the gender gap in reading and shrink it in math and science.”
In 20% of the countries, the gender gap completely disappeared or even switched after two hours of test duration.
Notably, Balart says girls’ superior ability to sustain performance “exists in all waves [of the PISA], and though the size and statistical significance varies somewhat across countries, given the results it is fair to say that this gender difference exists worldwide”.
He adds that he was surprised by “the robustness and pervasiveness of this gender difference”.
Other possible explanations recorded by the database – non-cognitive skills, test-taking strategies and test effort – did not influence the results.
To find out whether girls’ ability to sustain performance could help nullify perceived gender differences in maths, the researchers then analysed another database compiled for a meta-analysis of more than 400 maths tests.
By factoring in varying test length, they confirmed that girls did better on longer maths tests, closing the gender gap. As predicted, this was not explained by gender differences in dealing with time pressure.
Therefore, the “most notable implication of this study consists of emphasising a female strength in test taking,” says Balart, one that “has largely been ignored and that deserves visibility and recognition”.
“The research is yet another indication that cognitive tests do not only measure cognitive skills (how smart someone is), but that they also measure noncognitive skills (like perseverance and test motivation),” he adds.
He hopes the findings offer a counterbalance to gender stereotypes in maths and sciences that are reinforced by comments about dealing with time pressure, suggesting a female weakness.
Instead, the difference “could be framed in terms of rewarding a valuable skill in which female students perform relatively better”.
Although Balart cautions that test validity should be considered before changing their length, longer tests offer a valuable tool for policies seeking to achieve gender equality in STEM course enrolment and career choices.
