Wednesday, 11 May 2016

Why do women do so much better at university than their school test scores predict?

Picture an American high school staff-room, late in the academic year, where a teacher called Alice is listening to her colleagues ride their favourite hobby horse: picking out which students have the most promise.

Eventually Alice leans forward and taps her laptop. “Less talk, guys, more data. If you want to know how a student will do when they get to college, look at their aptitude test scores.” Betty throws her a look. “That won’t work,” she says, ”girls go on to do better than their test scores predict. Those tests are faulty.” Charles, the faculty provocateur, snorts. “Faulty? Not at all. Girls are only getting better grades because they pick softer subjects with easier marking.”

As her older colleagues tear into each other, Alice reflects on a third possibility: that succeeding at university depends on much more than the cognitive abilities measured by the SATs and ACTs (standard tests taken at American high schools), and that women might be better prepared in these other departments. But to resolve this staff-room squabble, who can tell these explanations apart?

Psychologists from the University of Minnesota-Twin Cities, that’s who. In a new paper in the Journal of Applied Psychology, Heidi Keiser’s team examine Alice and Charles’ rival explanations for why high school aptitude tests under-predict girls’ later success at university.

The new research first compared the university grades of 2000 students from a single institution with their high school aptitude scores. Women scored better on their course than you would expect based on an earlier aptitude test, but once the researchers took account of the female students’ higher average trait conscientiousness, 20 per cent of their grade surplus disappeared – a finding that replicates earlier research.

The researchers then decomposed the students’ degree course into elements, reasoning that if conscientiousness has a role in the gender gap, this should be greatest when grades depended highly on discretionary effort, like participating in discussion or research, and least when grades depended on raw smarts. The data showed that high school aptitude scores underestimated female performance on these effort-sensitive course elements, but were no worse at estimating their success on quizzes and tests than they were for men. Overall, this supports Alice’s perspective – that women do better than expected at university because of their greater effort and conscientiousness.

In a second study, the researchers tested Charles’ counter argument that women perform surprisingly well because they pick easier courses. The data, from huge historical datasets comprising nearly 400,000 students, showed that the courses men tended to take were significantly meaner (that is, male and female students on these courses tended to achieve worse grades than expected given their academic history) and were also more likely to be populated with high-achieving students competing for grades. And these factors, primarily course meanness, did explain a little of the overall tendency for female over-performance… but not more than nine per cent, much less than the effect of conscientiousness. A weak score, then, to Charles.

What about our other teacher Betty? Right from the start, she said the aptitude tests, measuring cognitive ability, simply weren’t doing it right. She could still have a case: the hidden variables found in this study – conscientiousness and course selection – together accounted for less than thirty per cent of the gender gap – possibly much less, if the two effects are not independent from each other.

However, it’s also plausible that aptitude tests are doing a reasonable job, it’s just that there are many non-cognitive factors critical to university success, and that conscientiousness is just one slice of this pie (an exploratory look at other personality variables by Keiser’s team suggests as much). If so, it’s not that school tests and exams need to be improved, but that they give us just one part of what higher education requires. We must not lose sight of the wider attributes, found particularly in female students, that travel from classrooms and school projects into our seminar rooms and lecture halls, and beyond.


Keiser, H., Sackett, P., Kuncel, N., & Brothen, T. (2016). Why women perform better in college than admission scores would predict: Exploring the roles of conscientiousness and course-taking patterns. Journal of Applied Psychology, 101 (4), 569-581 DOI: 10.1037/apl0000069

Post written by Alex Fradera (@alexfradera) for the BPS Research Digest.

Our free weekly email will keep you up-to-date with all the psychology research we digest: Sign up!

No comments:

Post a Comment

Note: only a member of this blog may post a comment.