When Stress Hits Hard, Students Say Mindfulness Works—But Only for Some People and Some Places

When people are told to manage their emotions, they often don’t do what they’re told—and now scientists have a clearer picture of why. A new study published in the journal Mindfulness by Almaz Mesghina of Northwestern University and Lindsey Engle Richland of the University of California, Irvine, asked nearly three hundred undergraduates to regulate their distress about the COVID-19 pandemic using either mindfulness or cognitive reappraisal, and then simply asked them whether they actually did it. The results expose a striking gap between the tidy world of emotion regulation theory and the messy reality of what people do when a stressor is personal, ongoing, and impossible to escape. Even more intriguing, the students’ own words about why their assigned strategy was or was not useful turned out to predict whether they complied at all, revealing a hidden layer of reasoning that laboratory experiments with button-press choices have largely overlooked.

The study built on a previous experiment conducted between October 2020 and March 2021, in the hardest stretch of the pandemic for college students. Participants, recruited from two large minority-serving institutions and predominantly identifying as Asian-American/Pacific Islander or Hispanic/Latinx, first reported their pandemic-related distress and trait anxiety. They then watched a three-minute instructional video for one of the two strategies, practiced it for another three minutes, and were told to apply it while watching a fourteen-minute educational lesson. The mindfulness video taught continual awareness of the present moment, nonjudgmental acceptance of emotions, and decentering from distressing thoughts, comparing thoughts to clouds drifting across the sky that one cannot touch or control. The reappraisal video, by contrast, framed stress itself as potentially adaptive—drawing on the Yerkes-Dodson inverted-U relationship between arousal and performance—and encouraged students to reinterpret their stress as the body recruiting the physiological and cognitive resources needed to meet a challenge.

After the lesson, each participant answered two questions: did they use the strategy, and why did they think it was or was not useful? Those open-ended answers became the raw material for an inductive thematic analysis conducted by three research assistants who were blind to both the experimental conditions and the study’s hypotheses. The coders developed data-driven themes rather than applying theory-driven categories, a deliberate choice meant to capture beliefs the researchers might never have anticipated. Half of all narratives were judged to express mixed or conditional efficacy—a proportion so large it reshaped the study’s focus. Trait anxiety and pandemic-related distress were equivalent across conditions and compliance groups, and narrative lengths did not differ, meaning the patterns the researchers found could not be explained away by who was more anxious, more distressed, or simply more verbose.

The headline finding was behavioral: mindfulness once again outperformed reappraisal in compliance, echoing a growing literature showing that acceptance-based strategies are used more often in daily life and impose fewer cognitive demands than reappraisal, which requires actively generating alternative interpretations of a stressor. But the qualitative data revealed something subtler. Participants who failed to use their assigned mindfulness practice were two to three times more likely than everyone else to describe contextual conditions as critical to its success, and four to six times more likely to invoke individual differences—claims that the strategy works for some brains and not others. Roughly 35 percent of mindfulness noncompliers pointed to context, often questioning whether the strategy fit remote learning environments at all. One student wrote that for mindfulness to work, one must be in an environment without noise; another argued it was better suited to in-person tasks, because online focusing resembles attending to a single object with many competing elements inside it.

The individual-differences narratives were the most consequential statistically. Controlling for condition, pandemic distress, and trait anxiety, mentioning that a strategy’s utility depended on who was using it predicted a 69 percent decrease in the odds of compliance—an odds ratio of 0.31, with a 95 percent confidence interval of 0.13 to 0.75. Some students framed this in terms of capacity: mindfulness, one wrote, is not useful for those who do not do well under pressure, because it will just wear them down. Others pointed to preexisting regulatory habits as a kind of sunk investment. One participant declined to change their approach because it was what had gotten them good grades so far; another noted that despite constant distraction in class they had never done badly, which had reinforced their refusal to change. As one student memorably put it, the strategy could not be universal because people think differently—like the saying that boiled water softens the potato but hardens the egg.

Perhaps the most counterintuitive result concerned people who did not comply yet still believed in the method. Nearly half of noncompliers’ narratives were coded as wholly or conditionally effective, and 60 percent of mindfulness noncompliers described their strategy as wholly effective—a higher proportion than among compliers in either condition. In other words, a person can recognize that a technique works, and even articulate its benefits, and still decline to use it. That finding directly challenges the assumption that perceived efficacy is the engine of compliance, and it supports a multi-step model of emotion regulation in which choosing a strategy, implementing it, and monitoring its success are separable decisions. Simply believing in a tool is evidently not the same as picking it up.

The reappraisal side of the study was harder to explain. Compliance was lower, but the thematic codes offered fewer clues about why noncompliant reappraisers chose differently from compliers. The researchers suggest that the pandemic, at the time of data collection, offered almost no reappraisal affordances: there was no plausible silver lining to a novel, uncontrollable, potentially deadly threat with sweeping academic and economic consequences. Without a viable alternative interpretation to construct, students assigned to reappraisal may have simply opted out of regulating altogether. Meanwhile, the finding that more distressed participants were more likely to comply with either strategy runs against earlier laboratory work suggesting that switching away from reappraisal increases with stressor intensity. The authors speculate that when a stressor is intense enough, any attempt at regulation—even a cognitively expensive one—can feel worthwhile.

The study has clear limitations, which the authors lay out with unusual candor. The analyses were exploratory and not preregistered, drawing on previously unanalyzed data from an earlier preregistered experiment. The binary compliance measure could not distinguish earnest full engagement from partial, delayed, or abandoned attempts, nor capture the iterative switching that researchers increasingly see as adaptive. Habitual use of mindfulness or reappraisal before the study was not measured, even though prior practice is known to shape both efficacy and compliance. Social desirability and obedience to instructions could also have inflated self-reported compliance. And because the sample consisted of highly distressed American undergraduates during an extraordinary global event, the findings may not generalize to other populations, calmer circumstances, or less self-relevant stressors. Factors known to matter—stressor controllability, cognitive effort, gender, age, and personality—were rarely mentioned spontaneously by participants, leaving open the question of whether these considerations are truly absent from everyday reasoning or merely unspoken.

Still, the implications reach well beyond pandemic-era coping. Mindfulness-based interventions suffer dropout rates ranging from a quarter to half of participants in apps and structured programs, yet when people do engage psychologically—not just physically—the reductions in distress can be substantial. This study suggests that a key to closing that gap lies in understanding the criteria people use when deciding whether a strategy fits them: their environment, their mental habits, and their sense of who they are. The authors recommend that all future emotion regulation research ask participants not only whether they complied, but whether they felt they needed to regulate in the first place, enabling more rigorous intent-to-treat analyses. In a field that has long treated strategy selection as the finish line, this work argues it is merely the starting gun—and that the reasons people give afterward, in their own words, may be the most honest data of all.

Subject of Research: Perceived utility of mindfulness versus reappraisal and their relation to emotion regulation compliance during real-life stress

Article Title: “Boiled Water Softens the Potato But Hardens the Egg”: Perceived Utility of Mindfulness and Reappraisal Strategies and Relations to Compliance In Situ

Article References: “Boiled Water Softens the Potato But Hardens the Egg”: Perceived Utility of Mindfulness and Reappraisal Strategies and Relations to Compliance In Situ. (n.d.). https://doi.org/10.1007/s12671-026-02995-4

Image Credits: AI Generated

DOI: 10.1007/s12671-026-02995-4

Keywords: emotion regulation, mindfulness, cognitive reappraisal, COVID-19, stress, compliance, college students, thematic analysis, acceptance, self-regulation, psychology, Boiled