Today’s Headscratch: Female Students Need More Exam Time?
I’ve been spending some time in this space working out my thoughts about some of the challenges presently facing higher education. (I reserve the right to change my mind, as always, as new evidence and perspectives are presented to me.)
Earlier, I posted my thoughts on educational reform and some thoughts on safety and free speech. Continuing on those themes are two recent articles in the news. The first is a piece in The Telegraph about how research and teaching content in universities is now more politicized than ever before, with ideology pushing evidence to the back seat.
The second, a glorious rant in The Chronicle of Higher Education, lists one professor’s frustrations with how universities have become “B.S.” He draws particular attention to how we sell more than teach, and again how ideology trumps genuine content and intellectual rigour.
I am shocked to find myself mostly in agreement with both positions. And I wonder if I’m just in a bitter time in my career, whether I’m being overly influenced by a handful of loud voices in the media, or if indeed there has been a geological shift in the fundament of the academy. I really don’t know. The echo chamber of social media sure doesn’t help my self-analysis.
Which leads me to another interesting development in the annals of (once again, British) higher education. At one of the greatest universities in the history of the world, Oxford, a computer science final exam was extended by 15 minutes —for everyone— because it was felt that “female candidates might be more likely to be adversely affected by time pressure.”
In other words, the time extension was put into place to help address the perceived deficit of women graduating with first class standing in their computer science program.
This is really quite interesting, and I struggle to understand the full justification and likely impact of the decision. Our first impulse is to examine the evidence that women are indeed more adversely affected by time pressures during exams, which I will do later in this post. But first let’s look at the the conceptual arguments in favour of this kind of accommodation.
I’ve been wrestling with this line of thinking for some time. I have an increasing proportion of students who, for a variety of medically defensible reasons, are granted additional time on exams, in excess of that given to the other students. The most common justification given is a diagnosis of anxiety. For such students, usually our university provides a special room where the student writes apart, and where a special proctor extends the time available. For my online class, in which exams are online and timed, I program the system to extend exam time for those accommodated students.
The idea behind accommodation is equality of opportunity. If you, for reasons beyond your control, are disadvantaged with respect to your ability to perform on the exam, then we will attempt to overcome that disadvantage to put you on equal footing with the other students.
The reason for this is that , according to the prevailing philosophy of modern higher education paedagogy, your performance in class is not an absolute, but rather is a relative standing. By this I mean that it means little if everyone gets A+. It matters more if I can separate the top performers from the poorer performers. So we accommodate the disadvantaged so that everyone starts at more-or-less an equal level, with only their performance on the exam being the prime indicator of relative class standing. In other words: our grading culture is based on in-group competition, not on any more abstract sense of individual progress, as is more prominent in elementary school.
In the Oxford case, all the students were given additional time, not just the presumed disadvantaged ones (i.e., women.) On its face, this is nonsensical, if one assumes that any time advantage experienced by the women will also be experienced by the men, thus raising everyone’s grades, but maintaining relative standing. And relative standing is ultimately what we care about, yes? In other words, if everyone is given the accommodation, then we need to dig deeper for a rationale to expect any change in relative standing.
I suppose it is possible that both men’s and women’s performance is asymptotic, with the female curve shifted a little positively on the x-axis. By this, I mean that after a certain time, it does not matter how much additional time is granted for the exam, since that additional time does not render a higher grade. (After a certain lengthy duration, everyone will just sit there, exam books closed, waiting for the time to expire.)
While both genders would have such a curve, the assumption that might be implicit in Oxford’s decision is that the drop-off in grade improvement for men, with respect to additional time, is steeper and/or happens earlier than that of women. In such a case, additional time would disproportionately raise women’s grades while not lowering men’s.
In this is the case, then I can see a viable argument for how this strategy is not unfair, but is merely accommodation. (Whether simply being of a certain gender is a “disability” or “shortcoming” needing accommodation is a touchier political question that I will not tackle here.)
However, it is also possible that this might affect relative standing in the course, if the accommodation experienced by women ends up being truly profound. I hope they are collecting data on this experience; the results will be paedagogically interesting.
In a grander philosophical sense, I often ask myself why I am examining my students at all. The easy first-order answer is that I am evaluating how well they can demonstrate what they have learned. The harder answer is that I am evaluating how well that can show what they’ve learned in a stressful scenario.
We can argue about whether there is utility in testing under stress. But if we accept that that is indeed what we are doing, then efforts to differentially reducing that stress for some students and not others is not so much accommodating a deficit, but rather conferring an advantage. And that is not fair.
An analogy would be a test of who can reach a basketball hoop. Some very tall students can do it without aid. Others would require a stool to accommodate their height deficit. But one could argue that the test is not “who can touch the hoop with a chair” but rather “who can touch the hoop. Period.” At that point, we have to step back and ask ourselves, who cares about whether someone can touch a hoop? What is the point of any kind of testing, and is the test valid? (Validity in science refers to whether the test is truly measuring the thing we claim it is.)
So the next level of analysis is, who cares if someone needs more time to finish an exam? All we should care about is whether or not they can do the job. Right?
I’ve had arguments with colleagues over this very question. If indeed timed exams are antiquated aberrations that should have no role in modern, enlightened education, then it should not matter that Oxford extended its exam time to help women. After all, shouldn’t the ultimate goal be to identify talented computer scientists, regardless of gender and regardless of their response to time pressures?
I don’t know the right answer to that question. Perhaps employers can offer some insight. Is it valuable to you to know that a potential employee can write an exam well under time pressure? Or are we just making up pointless hoops for students to jump through?
What we have not yet explored is the underlying assumption made by Oxford when adopting their exam strategy, specifically that timed exams differentially disadvantage women.
This article by Harvard scientists suggests that men are much more likely to select time-pressure math activities. When there is no time pressure, men and women performed the same on math-based challenges. But when there were time pressures, the men dramatically out-performed the women.
A similar South African study found that men and women show equal outcomes when there is no time pressure, but that there is a “strong negative impact on female’s performance” when there is a time constraint.
I am not a psychologist so will not pretend to understand the origins of this disparity. These data say nothing about whether the gender difference is innate (i.e. biological) or learned (i.e. socialized). Other studies of time pressure effects, such as this one at Huron College, failed to disentangle women from men.
If we assume that the assumption is true, that women do consistently and significantly respond more poorly to time pressures, and we accept that Oxford’s solution is a valid one, then what does this mean for other perceived gender effects in education?
If you’ve been paying attention you’ll know that there appears to be a crisis in education for men and boys at all ages, and in almost all subjects. In fact, a recent ground-shifting study suggests that this is not a Western phenomenon, but a global phenomenon. The authors also suggest that it might not even be recent, but has been with us for at least decades.
If we are prepared to re-assess exam timing strategies to address a dearth in women in computer sciences, what are we prepared to do to address a crisis in boys’ performance across all of public education?
Another study suggests that girls are truly only outperforming boys in “non-cognitive approaches to learning” (attentiveness, task persistence, eagerness to learn, learning independence, flexibility and organization), which then leads to better grades from teachers. This strikes me as a kind of Rosenthal or Pygmalion effect.
Whereas, male learning styles may be more competitive (hence the preference for timed tests), hierarchy based, and linked with physical activity. Increasingly, our learning and testing environments are becoming less competition- and hierarchy based and, the argument goes, less male friendly. Are we prepared to re-design the educational environment to address these presumptions?
But I digress. Again, I must point out that this is a blog, not a scientific journal. I’m just thinking aloud. I welcome your comments and criticisms, but please be respectful. Only through rational discourse can we find a way through the haze.