The research team, led by Simon Larose of Université Laval with colleagues from McGill University, the Université de Montréal, and Cégep de Saint-Laurent, followed a cohort of 1,348 students enrolled at ten colleges in Québec, Canada. Notably, 45 percent of the participants reported having a physical or psychological disability, a proportion high enough to allow meaningful comparisons between students with and without disabilities. The study’s design was unusually rigorous for pandemic-era research. Pre-pandemic academic adjustment and personal characteristics were measured in fall 2019, before anyone had heard of SARS-CoV-2, providing a genuine baseline. Stress and coping were then assessed at three points during the pandemic: spring 2020, spring 2021, and spring 2022. Finally, graduation data were drawn from official records in fall 2023, meaning the outcome measure did not rely on students’ self-reports.

Methodologically, the study employed a person-oriented approach rather than the variable-oriented analyses common in psychology. Instead of asking whether stress, on average, predicted graduation across the whole sample, the researchers used group-based trajectory modeling, a semiparametric technique developed by criminologist Daniel Nagin, to identify subgroups of students who followed similar developmental paths of stress and coping across the pandemic. This matters because averages can conceal divergent experiences: two students with identical mean stress levels might have followed entirely different courses, one climbing steadily and one spiking and recovering. By classifying students into trajectory groups, the team could ask which patterns of change were most common, which students ended up in which trajectories, and, crucially, whether trajectory membership predicted later graduation.

The first major finding concerned the shape of the pandemic’s psychological arc. Most students, the researchers found, experienced increasing stress between spring 2020 and spring 2021, followed by a subsequent decline after that point. This pattern tracks closely with the timeline of public health measures in Québec, where schools and colleges closed in March 2020, restrictions gradually eased from May 2021, and most sanitary measures were lifted by March 2022. The stress trajectory, in other words, mirrored the external constraints rather than simply fading with time. Meanwhile, the composition of students’ coping repertoires shifted in a more troubling direction. Over the same period, proactive coping strategies, the forward-looking efforts to anticipate problems and mobilize resources before crises hit, declined, whereas defensive coping strategies, the more reactive and avoidant responses aimed at managing distress after it arises, increased.

The second key finding concerned inequality in these trajectories. Students with disabilities were disproportionately represented in the trajectory groups characterized by high stress and high defensive coping, and they were underrepresented in the groups marked by high proactive coping. This is not merely a statistical curiosity. Prior research, including the team’s own earlier work on the first pandemic wave, had suggested that students with disabilities faced distinctive stressors during COVID-19, from disrupted access services to the cognitive and sensory demands of remote learning. The new analysis extends that picture across time, showing that the pandemic’s psychological burden was not evenly distributed even as conditions evolved. A student who entered college with a disability was more likely to ride the high-stress, defensively-coping current of the pandemic years, and less likely to be swimming with the proactive copers.

The third finding is the one with the sharpest practical edge. After controlling for pre-pandemic academic adjustment and other personal characteristics measured in fall 2019, membership in a high defensive coping trajectory was associated with lower odds of graduating by fall 2023, regardless of disability status. The wording of that last clause deserves emphasis. The link between defensive coping and non-graduation held for students with and without disabilities alike, suggesting that the mechanism is not specific to disability but rather reflects a general vulnerability conferred by a maladaptive coping style sustained across the crisis. In effect, students whose response to pandemic stress hardened into defensive patterns paid an academic price years later, even after the researchers statistically accounted for how well adjusted they had been before the pandemic began.

Why would defensive coping erode graduation odds? The authors’ framework, grounded in stress and coping theory, offers a plausible account. Proactive coping is resource-building: it involves planning, seeking support, and acting early to prevent stressors from escalating. Defensive coping, by contrast, is resource-draining: it manages immediate distress through avoidance, denial, or disengagement, which may provide short-term relief while allowing academic problems, missed deadlines, skipped courses, mounting incompletes, to accumulate silently. Over a multi-year program, that accumulation can become decisive. A student who repeatedly disengages from a difficult course in the moment may find themselves off-track for graduation long before the problem announces itself. The pandemic, by sustaining stress at elevated levels for two years, may have given defensive coping unusually fertile ground in which to take root and entrench.

The study’s implications for colleges and universities are relatively direct. If defensive coping trajectories predict dropout, then interventions that promote adaptive, proactive coping strategies could support academic persistence and reduce dropout risk among students with and without disabilities. The authors highlight the importance of such promotion explicitly in their conclusions. This aligns with a growing evidence base on student mental health interventions, including randomized trials of online mindfulness programs and peer support programs delivered during the pandemic, which have shown that psychological skills can be taught at scale. The present findings suggest such efforts should not be framed solely as mental health initiatives but as academic persistence initiatives, with graduation itself as an outcome worth targeting. They also suggest that students with disabilities warrant particular attention, given their overrepresentation in the high-risk trajectories, though the benefit of fostering proactive coping appears to extend to the entire student body.

Several caveats temper the conclusions. The cohort was drawn from Québec colleges, institutions with their own structural features, and graduation data came from that provincial system, so replication elsewhere is needed before generalizing broadly. Stress and coping were self-reported at annual intervals, which cannot capture week-to-week fluctuations or the mechanisms linking coping to academic outcomes. And while the statistical controls for pre-pandemic adjustment strengthen causal interpretation, they cannot fully rule out the possibility that unmeasured characteristics influence both coping trajectories and graduation. Still, the design’s strengths, a true pre-pandemic baseline, official graduation records, a large disability-inclusive sample, and trajectory methods that respect individual differences, make this one of the more credible longitudinal accounts of the pandemic’s academic toll. As institutions continue to absorb the aftershocks of COVID-19, the study’s core lesson stands out: the students most at risk of not graduating may be identifiable not by their stress levels alone, but by the defensive shape of their coping, and that shape, unlike a disability or a pandemic, is something colleges can help change.

Subject of Research: Longitudinal associations between stress and coping trajectories during COVID-19 and college graduation in students with and without disabilities

Article Title: College Graduation in Students with and without Disability: Links with Stress and Coping Trajectories during COVID-19

Article References: Larose, S., Beaulieu, C., Aziz, C., Bationo, N. J.-C., Bureau, J. S., Geoffroy, M.-C., Ghetti, A., & Côté, S. (2026). College Graduation in Students with and without Disability: Links with Stress and Coping Trajectories during COVID-19. Higher Education. https://doi.org/10.1007/s10734-026-01783-2

Image Credits: AI Generated

DOI: 10.1007/s10734-026-01783-2

Keywords: COVID-19, college graduation, stress, coping strategies, students with disabilities, higher education, longitudinal study, defensive coping, proactive coping, academic persistence, dropout risk, Québec colleges

The research, examined in a News & Views commentary by Bryce E. Hughes of Montana State University, reframes the gender gap not simply as a matter of who enters which field, but as a matter of who is allowed to struggle, who is carried through, and who is held to unforgiving standards. According to Hughes, Cimpian and his team demonstrate that the same polarization that has served as a barrier to gender-nonconforming students simultaneously functions as a protective buffer for conforming students with weaker academic records. In other words, fitting the gendered stereotype of a field can buy a student leniency, while deviating from it can cost a student dearly, regardless of ability.

The historical backdrop of this work is well documented. Claudia Goldin, Lawrence Katz and Iyeko Kuziemko famously traced the twentieth-century reversal in which women came to outnumber men in American colleges, transforming the demographics of higher education within a single generation. Yet the horizontal segregation of the curriculum persisted. Women moved into higher education in record numbers but remained concentrated in particular majors, while men clustered in others. That horizontal divide, measured in fields such as engineering and nursing, has proven remarkably durable, and researchers have long treated it as the central mechanism through which gender inequality is reproduced in universities.

What the new study makes clear is that this divide operates through more than admission gates and chilly classroom climates. Gender polarization, in the sense used by the researchers, describes the extent to which a field is culturally coded as masculine or feminine, and the degree to which students are expected to conform to the gender norms associated with their chosen discipline. When a field is strongly polarized, conformity is rewarded and deviation is penalized. The penalty has been the traditional focus of research on women in engineering and men in nursing. The reward, the new findings suggest, has gone largely unnoticed: students whose gender matches the field’s dominant image receive a kind of implicit sponsorship that can carry them through academic difficulty.

The technical heart of the analysis lies in how the researchers connect field-level gender composition and gender-typing to individual student outcomes. Rather than treating all students in a male-dominated field as facing identical conditions, the study disaggregates outcomes by gender conformity and by academic performance. The pattern that emerges is one of differential tolerance. Lower-performing students who conform to a field’s gender norms are more likely to persist and complete their degrees than their performance alone would predict, as though the field’s cultural expectations extend a presumption of belonging that overrides evidence of struggling. Students who do not conform receive no such presumption, and their academic difficulties are correspondingly more consequential.

The implications for LGBTQ+ students are among the most consequential findings reported. Hughes notes that Cimpian and colleagues show that polarization impedes LGBTQ+ students specifically in male-dominated fields, compounding the disadvantages already documented for sexual and gender minorities in science, technology, engineering and mathematics. Previous work by Hughes himself, published in Science Advances in 2018, found that sexual minority students in STEM majors were substantially less likely to complete their degrees than their heterosexual peers, a gap that could not be explained by prior achievement alone. The new study provides a structural explanation for why that gap persists: in fields where masculinity is tightly bound to professional identity, students who are perceived as deviating from heterosexual masculine norms lose access to the protective sponsorship that shields their conforming classmates.

This mechanism connects to a broader theoretical literature on status beliefs and expectation states. Research by Shelley Correll and Cecilia Ridgeway has described how shared cultural beliefs about gender function as status systems, distributing deference, benefit of the doubt and evaluative leniency along gendered lines even when formal rules are neutral. Moss-Racusin and colleagues have documented how these beliefs shape judgments of competence in applied settings, and work by Doyle and Molix has explored how stigma related to sexual orientation undermines belonging and performance through social psychological pathways. The new findings translate these abstract status dynamics into concrete educational outcomes: the benefit of the doubt, quantified, shows up as persistence and completion rates.

The study also resonates with long-standing questions about how academic merit is defined and measured. Alexander Astin’s classic work on educational excellence argued that institutions routinely confuse selectivity with quality, rewarding inputs rather than development. The new research extends that critique into the gendered dimension of evaluation. If conforming students are buffered from the consequences of weak performance while nonconforming students are not, then the academic records that emerge from polarized fields are not clean measures of ability. They are products of a system in which gender conformity functions as an unearned academic asset, inflating the measured success of some students while exposing others to full accountability for every stumble.

For policymakers and educators, the findings suggest that interventions aimed solely at recruitment are unlikely to close gender gaps in fields such as engineering. Bringing more women, gender-nonconforming students and LGBTQ+ students into a polarized field does not dismantle the conformity economy that operates inside it. If the field continues to extend leniency to those who fit its dominant image and scrutiny to those who do not, the composition of its graduates may change slowly while its internal inequities remain intact. Hughes suggests that the research points instead toward examining the evaluative cultures of disciplines themselves: how belonging is granted, how struggle is interpreted, and whose difficulties are read as fixable rather than disqualifying.

The study ultimately asks a question that extends well beyond campus walls. Gender polarization in the labor force, in professional licensing and in occupational prestige hierarchies operates on the same logic that Cimpian and colleagues have now documented in higher education. If conformity to gender norms buys protection from failure, then the gender gap is not merely a gap in opportunity but a gap in forgiveness, and closing it will require dismantling the invisible sponsorship that has quietly sustained it. As Hughes concludes in his commentary, reconsidering the gender gap means recognizing that the barrier and the shield are two faces of the same structure, and that neither can be addressed without confronting the other.

Subject of Research: How gender polarization in academic fields affects student persistence and outcomes by gender conformity and sexual orientation

Article Title: The gender gap reconsidered

Article References: Hughes, B. E. (2026). The gender gap reconsidered. Nature Human Behaviour. https://doi.org/10.1038/s41562-026-02448-x

Image Credits: AI Generated

DOI: 10.1038/s41562-026-02448-x

Keywords: gender gap, higher education, STEM, LGBTQ+ students, gender polarization, academic persistence, engineering, nursing, Nature Human Behaviour, educational equity, gender norms, student outcomes