The sample included 69 adults who self-identified as people who stutter and 71 who did not, with participants aged 18 to 65 years. Stuttering status was based on self-identification without independent clinical confirmation. Participants completed the Cognitive Emotion Regulation Questionnaire and the Protective Factors for Resilience Scale. The researchers compared sex, age and education between groups, used independent-samples t tests or Mann–Whitney U tests for questionnaire outcomes, and calculated effect sizes with 95% confidence intervals. Within-group relationships were examined with Pearson or Spearman correlations, with Benjamini–Hochberg false discovery rate correction applied separately to 36 subscale-level correlations in each group.

The groups did not differ significantly in sex distribution, age or education. No statistically significant between-group differences were detected for any of the nine CERQ subscales, and the corresponding effect sizes were small; the abstract does not establish equivalence between the groups. People who stutter reported lower Personal, Peer, Family and Total Protective Factors for Resilience Scale scores, with small-to-moderate rank-biserial effects. After false discovery rate correction, 12 positive associations involving positive refocusing, refocus on planning and positive reappraisal remained statistically significant within the group reporting stuttering. No correlations remained significant in the group not reporting stuttering.

These results indicate that the study detected lower reported resilience-related protective resource scores among participants who stutter, alongside no detected differences in measured cognitive emotion regulation strategies. The design and analysis do not demonstrate that resilience resources cause emotion regulation strategies, that stuttering causes lower scores, or that the two constructs have different correlation strengths across groups. The cross-sectional design, exploratory analysis, modest sample size and reliance on self-identification of stuttering status limit interpretation. The available source provides the study abstract rather than the complete article, so this report is limited to the findings and methods stated there.

Subject of Research: Cross-sectional comparison of cognitive emotion regulation and resilience-related protective factors in adults who stutter

Article Title: Cognitive emotion regulation strategies and resilience-related protective factors in persons who stutter: a cross-sectional comparative and correlational study

Article References: Alpüran Kocabıyık, N., Erol, E., & Sarı Sülüker, A. (2026). Cognitive emotion regulation strategies and resilience-related protective factors in persons who stutter: a cross-sectional comparative and correlational study. BMC Psychology. https://doi.org/10.1186/s40359-026-05815-5

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05815-5

Keywords: stuttering, adults, emotion regulation, resilience, protective factors, cross-sectional study, psychology, questionnaire, correlation, Cognitive, emotion, regulation

The research team, led by U. Venkatesh of the All India Institute of Medical Sciences in Gorakhpur, together with colleagues from AIIMS campuses in Nagpur, Rajkot, Deoghar and Guwahati as well as JIPMER in Puducherry, set out to address a persistent gap in the literature. Most research on adolescent substance use has concentrated on risk factors such as peer pressure, family conflict and easy access to psychoactive substances. Far less attention has been paid to the psychosocial assets that may buffer young people against initiation and continued use, and the evidence that does exist comes disproportionately from high-income countries. For low- and middle-income settings, where the majority of the world’s adolescents live, robust data on protective resources have been scarce.

The study defined current substance use as the self-reported use of at least one assessed psychoactive substance during the preceding three months. Complete information on this outcome was available for 4,848 of the 6,168 enrolled adolescents, a response coverage of 78.6 percent. Within this analytic sample, 545 adolescents, or 11.2 percent, reported current substance use. Rather than singling out one substance, the researchers cast a wide net across the psychoactive substances commonly assessed in adolescent health surveys, capturing the full spectrum of use from tobacco and alcohol to other substances.

Methodologically, the investigators constructed study-specific composite scores for three domains: protective factors, coping mechanisms, and social engagement. Each score was assembled from questionnaire items embedded in the parent multicentre study and then categorized as low, moderate, or high using cut-offs derived from the interquartile range of the score distributions. This approach allowed the team to examine whether associations with substance use differed across the spectrum of psychosocial resources rather than treating these constructs as continuous variables, an important design choice given that the relationship between psychological resources and behavior is not always linear.

The headline result concerned protective factors. Compared with adolescents in the low category, those with moderate protective-factor scores had 40 percent lower odds of current substance use, an adjusted odds ratio of 0.60 with a 95 percent confidence interval of 0.47 to 0.76, which excludes the null value and therefore reaches conventional statistical significance. Strikingly, the estimate for adolescents with high protective-factor scores was weaker and less precise, at 0.78 with a confidence interval of 0.59 to 1.02, a range that includes the null. The same pattern emerged for coping mechanisms: moderate coping scores were associated with 35 percent lower odds of current use, an adjusted odds ratio of 0.65 with a confidence interval of 0.51 to 0.82, while high coping scores yielded a weaker estimate of 0.81 with a confidence interval of 0.60 to 1.10 that crosses the null.

Social engagement told a different and more equivocal story. The moderate category showed essentially no independent association with substance use, with an adjusted odds ratio of 0.98 and a confidence interval spanning 0.77 to 1.25. The high-category estimate, at 2.27 with a very wide confidence interval of 0.71 to 7.24, was highly imprecise and cannot support any firm conclusion about whether greater social engagement raises, lowers, or leaves unchanged the odds of adolescent substance use. The researchers also conducted exploratory interaction analyses to test whether peer norms modified these relationships, finding only limited evidence of such effect modification.

To probe the overall direction of the associations, the team performed adjusted ordinal analyses examining trends across increasing categories of each psychosocial measure. These analyses indicated overall inverse ordinal trends, meaning that, on average, higher categories of protective factors and coping were associated with lower substance use. However, the authors are emphatic that the category-specific estimates did not consistently demonstrate monotonic patterns. Because the moderate categories showed stronger associations than the high categories, the findings should not be interpreted as evidence of dose-response relationships, in which each incremental rise in psychosocial resources produces a corresponding drop in substance use. This distinction matters both for statistical interpretation and for the practical design of interventions.

The statistical models were carefully adjusted for a battery of potential confounders, including age, sex, household economic status, living arrangement, the amount of pocket money available to the adolescent, peer norms, and study site. This adjustment is critical because older adolescents, boys, those with more disposable income, and those in permissive peer environments are known to differ systematically in both psychosocial resources and substance use. By holding these factors constant, the analysis isolates the association between the psychosocial composites and substance use, though it cannot eliminate confounding by unmeasured variables.

The authors are equally candid about the limits of what a cross-sectional design can establish. Because exposure and outcome were measured at the same point in time, the study cannot determine whether lower protective factors and coping capacity preceded substance use or, conversely, whether substance use eroded those resources. Temporality and causality remain unresolved questions. The composite scores, while constructed systematically from study questionnaire items, were not validated instruments, which introduces the possibility of measurement error. The authors call explicitly for longitudinal studies using validated psychosocial measures to clarify these associations over time.

Even with these caveats, the study carries significant implications for adolescent health programming in India and comparable settings. The work is embedded in a larger research effort funded by the Indian Council of Medical Research under an Extramural Research Grant, part of a cluster randomized controlled trial evaluating a community-based peer-led intervention to prevent substance use among adolescents within the school health programme under Ayushman Bharat, India’s national health initiative. If subsequent longitudinal research confirms that moderate psychosocial resources confer the greatest protection, school-based programmes may need to look beyond simply maximizing every student’s resilience scores and instead focus on ensuring that no adolescent falls below a functional threshold of protective resources. The ethical scaffolding of the study was rigorous: approvals were obtained from the institutional ethics committees of all six participating institutions, informed consent was secured from all adolescents, written parental or guardian consent was obtained for participants under 18, participation was voluntary, and confidentiality was strictly maintained, in accordance with the Declaration of Helsinki. As substance use during adolescence remains a pressing public health concern with consequences that ripple across the life course, this large, multi-regional analysis offers policymakers a data-driven starting point, and a caution against linear thinking, in designing prevention strategies for the world’s largest adolescent population.

Subject of Research: Psychosocial protective factors, coping mechanisms and social engagement associated with current substance use among school-going adolescents in India

Article Title: Protective factors, coping mechanisms and social engagement associated with current substance use among adolescents from 108 schools in India: a multicentre cross-sectional study

Article References: Venkatesh, U., Gandhi, A. P., Joshi, K., Bhattacharya, S., Agarwalla, R., Chinnakali, P., & Hossain, M. (2026). Protective factors, coping mechanisms and social engagement associated with current substance use among adolescents from 108 schools in India: a multicentre cross-sectional study. BMC Psychology. https://doi.org/10.1186/s40359-026-05658-0

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05658-0

Keywords: adolescents, substance use, protective factors, coping mechanisms, social engagement, India, school health, psychosocial resources, cross-sectional study, public health, BMC Psychology, prevention

Researchers at the Lifespan Brain Institute of the Children’s Hospital of Philadelphia and the University of Pennsylvania took advantage of a fortuitous scientific circumstance. Years before the pandemic began, between May 2017 and August 2019, they had administered in-person clinical assessments and resilience questionnaires to a cohort of youth and young adults as part of work developing the Lifespan Risk and Resilience Battery. When the pandemic struck, the team recontacted participants and assessed their mental health remotely between May and August 2020, roughly 21 months after the baseline visits. Because resilience factors were measured before anyone knew a global crisis was coming, the design sidesteps the recall and reverse-causation problems that plague most pandemic-era research.

The analytic sample included 96 participants with a mean baseline age of 19.34 years, ranging from 8 to 29 years old. Fifty-one percent were female and 59 percent identified as People of Color. Critically, the cohort was psychiatrically enriched: 70 percent met criteria for a lifetime or current psychiatric disorder, including schizophrenia spectrum disorders in 15 percent and subthreshold psychosis-spectrum syndromes such as attenuated psychosis syndrome in 36 percent. Thirty percent had no diagnosis. This deliberate clinical heterogeneity meant the researchers could ask whether pre-existing resilience operated across a broad spectrum of psychiatric vulnerability, not merely in healthy community samples where the stakes are lower.

The study focused on two theoretically grounded resilience processes drawn from the 47-item Risk and Resilience Battery. The first was emotion regulation, assessed with five items from the Difficulties in Emotion Regulation Scale Short Form that capture difficulties modulating emotional responses, such as trouble concentrating or controlling behavior when upset. Scores were reverse coded so higher values reflected greater regulatory capacity. The second was supportive close relationships, measured with four items from the Network of Relationships Inventory assessing the perceived quality of close family bonds, including feelings of care, admiration, and stability with parental figures and siblings. These factors were chosen because developmental resilience frameworks, notably Ann Masten’s account of resilience as ordinary magic arising from interacting intrapersonal and interpersonal systems, position them as distinct yet interconnected regulatory resources.

At follow-up, participants completed brief self-report measures of depression, anxiety, psychosis-spectrum symptoms, and sleep disturbance, including the Patient Health Questionnaire-2, the Generalized Anxiety Disorder-7, a five-item PRIME short form, and the Insomnia Severity Index. The analytical strategy was rigorous. Multivariate regression models controlled for age, sex, race, baseline psychotropic medication use, time between visits, and, crucially, corresponding baseline symptom levels, so the findings reflect protection against new or worsening symptoms rather than simple symptom stability. Given the moderate sample size, the team used nonparametric bootstrapping with 5,000 resamples to generate bias-corrected and accelerated 95 percent confidence intervals, treating coefficients as significant only when those intervals excluded zero.

The results were clear and, in places, striking. Higher pre-pandemic emotion regulation predicted lower anxiety during the pandemic (p = .007), fewer psychosis-spectrum symptoms (p = .034), and fewer sleep problems (p = .002). Notably, resilience factors added 6.7 percent of explained variance to the anxiety model beyond covariates and baseline anxiety, a substantial incremental contribution. Supportive close relationships independently predicted fewer sleep problems (p = .022), although they were not associated with depression, anxiety, or psychotic-like experiences. Neither resilience factor related to depressive symptoms, and neither significantly changed the odds of crossing the screening threshold for elevated depression in supplemental logistic analyses. Baseline psychotropic medication use was not a significant predictor in any model.

The transdiagnostic breadth of the emotion regulation finding carries particular weight. Emotion regulation difficulties have long been implicated across depression, anxiety, and psychosis-spectrum conditions, and stress-diathesis models of schizophrenia posit that stress sensitivity and poor coping fuel psychotic-like experiences such as unusual thoughts, suspiciousness, and perceptual disturbances. The new data extend this literature by showing that regulatory capacity measured nearly two years before a chronic stressor predicted symptom levels across affective and psychosis-spectrum domains while that stressor was still unfolding. Mechanistically, the authors argue, emotion regulation likely functions as an upstream regulator of stress reactivity: individuals with stronger regulatory skills can modulate cognitive, affective, and physiological responses to adversity, reducing vulnerability that would otherwise cascade into multiple symptom domains.

The protective link between supportive family relationships and sleep points to a complementary, downstream mechanism. Theoretical accounts of interpersonal co-regulation propose that close relationships stabilize physiological arousal, dampen hypervigilance, and help restore daily rhythms, processes that are especially important for sleep onset and maintenance. The social buffering literature similarly suggests that supportive ties attenuate hypothalamic-pituitary-adrenal axis reactivity and autonomic arousal, both of which disrupt sleep. During lockdowns, when families became the primary source of co-regulatory support, relationships characterized by emotional reassurance and felt security may have helped preserve healthy sleep routines. Because sleep disturbance is itself a transdiagnostic risk factor for mood, anxiety, and psychosis-spectrum conditions, protecting sleep during chronic stress may prevent broader psychiatric cascades.

The authors are careful to note that these two resilience processes probably do not operate in isolation. Supportive relationships may scaffold emotion regulation capacities through co-regulatory exchange, while individual regulatory skills in turn shape relationship quality and stability over time, creating reciprocal feedback between intrapersonal and interpersonal systems. The lifespan perspective of the sample, spanning ages 8 to 29, captures a window in which prefrontal regulatory networks continue maturing and the primary sources of support shift from parents toward peers, although the study was underpowered to formally test developmental moderation. Limitations also include the absence of baseline sleep measurement, differing symptom instruments across waves, unknown medication status at follow-up, and a clinically enriched sample that may limit generalizability to fully community-based populations.

Even with those caveats, the implications are consequential. As pandemics, climate-related disasters, and political instability make large-scale stressors an increasingly recurring feature of modern life, the study argues for investing in resilience before crises arrive rather than scrambling to treat distress afterward. Emotion regulation is demonstrably amenable to intervention: school-based social-emotional learning programs and transdiagnostic cognitive-behavioral and mindfulness-based therapies have been shown to strengthen regulatory skills across diverse populations. Supportive relationships, meanwhile, can be cultivated through family-level interventions that enhance emotional availability and co-regulation. By showing that qualities measured in ordinary times predicted who fared better during extraordinary ones, the Philadelphia team has provided some of the strongest prospective evidence yet that resilience is not a vague abstraction but a set of identifiable, modifiable, and measurable protective processes with the potential to buffer mental health when the next global storm arrives.

Subject of Research: Prospective prediction of youth mental health during the COVID-19 pandemic from pre-pandemic emotion regulation and supportive close relationships

Article Title: Pre-Pandemic Emotion Regulation and Supportive Relationships as Protective Factors Across Development: Prospective Prediction of Mental Health During the COVID-19 Pandemic

Article References: Kim, I. A., Ered, A., Himes, M. M., Gur, R. E., Rush, S., Moore, T. M., Barzilay, R., Calkins, M. E., & White, L. K. (2026). Pre-Pandemic Emotion Regulation and Supportive Relationships as Protective Factors Across Development: Prospective Prediction of Mental Health During the COVID-19 Pandemic. Child Psychiatry & Human Development. https://doi.org/10.1007/s10578-026-02086-4

Image Credits: AI Generated

DOI: 10.1007/s10578-026-02086-4

Keywords: emotion regulation, resilience, COVID-19 pandemic, mental health, supportive relationships, sleep problems, anxiety, psychosis-spectrum symptoms, youth development, protective factors, longitudinal study, Child Psychiatry & Human Development

The study, conducted by Paweł Holas, Maria Anna Wasylkowska, Piotr Grajewski, and Patryk Roczon at the University of Warsaw, surveyed 763 adolescents aged 11 to 18, with a mean age of 15.27 years. Roughly half of the participants were female. Using validated self-report instruments administered in classrooms during the school day, the team estimated how common problematic internet use had become among Polish youth and mapped the psychological terrain that surrounds it. Roughly 5.9 percent of the sample, 45 adolescents, met the criteria for problematic internet use, a figure that closely mirrors global prevalence estimates of around six percent.

Problematic internet use is not yet a formal psychiatric diagnosis, but it draws heavily on the logic of behavioral addiction. Clinicians and researchers describe it as excessive, poorly controlled online behavior that damages personal, social, or academic functioning. It shares hallmark features with substance and behavioral addictions, including mood modification, tolerance, salience, and withdrawal-like distress when access is cut off. Adolescents are considered especially vulnerable because the neural circuitry that governs self-regulation and impulse control is still maturing, making compulsive patterns of use easier to establish and harder to break.

The health stakes are considerable. Prior research links problematic internet use to disturbed sleep, poor posture, obesity, and visual strain, as well as to depression, anxiety, stress, and attention-deficit/hyperactivity disorder. It also co-occurs with substance use, including cannabis, nicotine, and alcohol, and with behavioral addictions such as gaming disorder. Among teenagers specifically, it is associated with deteriorating peer relationships and declining school performance. Because these harms span so many domains, identifying modifiable risk and protective factors has become a central goal of prevention science.

The Polish team measured a broad battery of psychological variables: problematic internet use severity with a short form of Young’s Internet Addiction Test, generalized anxiety with the GAD-7, depressive symptoms with the PHQ-2D, perceived stress with the PSS-4, psychological inflexibility with the Avoidance and Fusion Questionnaire for Youth, emotional dysregulation with the short form of the Difficulties in Emotion Regulation Scale, and trait mindfulness with the Child and Adolescent Mindfulness Measure, recently validated in Polish adolescent samples. All scales showed acceptable to excellent internal consistency, with reliabilities ranging from 0.75 to 0.91, with the exception of the two-item depression screener, which showed only low-to-moderate reliability.

The statistical results formed a coherent picture. In hierarchical regression models, greater emotional dysregulation predicted higher problematic internet use severity, while higher trait mindfulness and the presence of siblings each predicted lower severity. Higher psychological inflexibility, anxiety, depression, and stress were each positively associated with problematic use in bivariate correlations. The final regression model accounted for 21.6 percent of the variance in problematic internet use severity, a substantial share for psychological predictors in this field.

The most technically interesting part of the study is its mediation analysis, estimated with structural equation modeling and 10,000 bootstrap resamples. The researchers positioned trait mindfulness as an upstream correlate and tested whether depressive symptoms and generalized anxiety symptoms carried its association with problematic internet use. Depressive symptoms emerged as a significant partial mediator: lower mindfulness was associated with higher depressive symptoms, which in turn were associated with greater problematic use severity, producing a significant indirect effect with a bootstrap confidence interval that excluded zero. Meanwhile, the direct negative association between mindfulness and problematic use remained significant, indicating that depression explains only part of the pathway. Generalized anxiety symptoms, by contrast, did not significantly mediate the relationship, even though mindfulness was strongly inversely associated with anxiety itself.

Group comparisons reinforced these findings. Adolescents classified as having problematic internet use scored significantly worse than their peers on anxiety, psychological inflexibility, emotional dysregulation, and perceived stress, and significantly lower on mindfulness. This pattern fits theoretical models, including acceptance and commitment therapy frameworks and the integrative I-PACE model of addictive behaviors, in which problematic online behavior functions as a maladaptive coping strategy aimed at escaping or numbing negative affect. Within that logic, mindfulness operates as a foundational capacity: adolescents who can observe difficult emotions without immediately reacting may have less need to retreat into the screen.

One of the more unexpected findings concerns family structure. The presence of siblings was associated with lower problematic internet use severity, a novel observation in a Polish sample that echoes earlier longitudinal work suggesting sibling relationships provide social support and help children regulate emotions. The authors suggest that fostering positive family dynamics and social connectedness could complement mindfulness-based strategies in prevention programs, although this correlational finding requires replication before strong claims are warranted.

The researchers are careful about what the data can and cannot show. Because the design is cross-sectional, the associations are correlational and cannot establish that mindfulness causally reduces problematic internet use; alternative directional models, in which distress erodes mindful awareness or in which excessive use itself dampens present-moment attention, remain plausible and would require longitudinal or experimental designs to test. Self-report measures are also vulnerable to reporting bias, and the low reliability of the two-item depression measure may have added noise to the mediation estimates. Still, the study aligns with a growing evidence base: meta-analyses show that mindfulness-based programs significantly reduce problematic internet use symptoms, and longitudinal studies link increases in trait mindfulness to declines in problematic use over time. Taken together, the findings position dispositional mindfulness as a promising candidate protective factor, and depressive symptoms as a key mechanism worth targeting, in the continuing effort to keep adolescents grounded in a hyperconnected world.

Subject of Research: Trait mindfulness as a protective factor against adolescent problematic internet use, with depressive symptoms as a mediator

Article Title: Adolescent problematic internet use: protective associations of trait mindfulness and the mediating role of depressive symptoms

Article References: Holas, P., Wasylkowska, M. A., Grajewski, P., & Roczon, P. (2026). Adolescent problematic internet use: protective associations of trait mindfulness and the mediating role of depressive symptoms. Current Psychology, 45(17), Article 1491. https://doi.org/10.1007/s12144-026-10048-1

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10048-1

Keywords: problematic internet use, mindfulness, adolescents, depressive symptoms, anxiety, emotional dysregulation, psychological inflexibility, mediation analysis, behavioral addiction, mental health, protective factors, Current Psychology

The study, published in the journal Large-scale Assessments in Education, was conducted by Markéta Žáková and Tomáš Lintner of Masaryk University in the Czech Republic. Drawing on PISA 2018 data from more than 600,000 fifteen-year-old students across 75 education systems, the pair set out to answer a deceptively simple question: does it matter which operational definition of academic resilience a researcher uses? The answer, they found, is a resounding yes, with consequences that ripple through prevalence estimates, country rankings, and conclusions about which factors help disadvantaged students beat the odds.

Academic resilience combines two ingredients: adversity, usually measured by socioeconomic status, and positive adaptation, usually measured by achievement. But researchers disagree about how to stitch those ingredients together. One common approach simply identifies students who land in the bottom slice of the socioeconomic distribution and the top slice of achievement, for example the poorest quarter who score in the top quarter. A second approach instead computes what each student’s achievement should be, statistically speaking, given their family background, and flags students who substantially outperform that expectation, an approach known as the residual method. A third treats resilience as a process rather than an outcome, predicting achievement as a continuous variable within the disadvantaged group and never labeling anyone resilient at all.

Žáková and Lintner compared these approaches systematically. They estimated a top-achiever definition, a residual definition computed within each country, and a residual definition benchmarked against an international standard, each at three different threshold levels, 20, 25, and 33 percent, spanning the range most commonly used in the published literature. All told, this produced nine binary operationalizations plus the continuous-outcome model, yielding nearly a thousand country-level analyses. Each estimate was pooled across ten plausible values for achievement and twenty multiply imputed datasets, with standard errors that fully accounted for PISA’s complex two-stage sampling design.

The headline finding is stark. The cross-country mean prevalence of academically resilient students ranged from 7.5 percent under the strictest definition to 35.3 percent under the most inclusive residual benchmark, a nearly fivefold difference computed from identical data. Part of this spread is arithmetic, since looser thresholds mechanically admit more students, but the deeper problem emerges when countries are ranked. Rankings were reasonably stable across thresholds within a given definition, yet they diverged sharply across definitions. The correlation between rankings produced by the top-achiever approach and the residual approaches fell as low as 0.51, and between the two residual variants, which differ only in whether the statistical expectation is local or global, it dropped to between 0.32 and 0.47. A country celebrated as an equity champion under one definition can rank unremarkably under another, and the discrepancy is itself informative about what its disadvantaged students actually do well.

The authors illustrate why with a thought experiment grounded in their results. A lower-performing system with a steep socioeconomic gradient may rank poorly on the top-achiever definition, because few of its disadvantaged students reach absolute excellence, yet rank highly on the within-country residual definition, because modest local expectations are easy to exceed. That same system may then sink again on the internationally benchmarked residual definition, since beating a weak local bar is not the same as meeting a global standard. The researchers argue that resilience rankings should therefore be reported under multiple definitions side by side, as complementary views of equity rather than competing estimates of a single quantity, a practice the OECD itself briefly adopted nearly a decade ago.

What about the protective factors that resilience research is meant to uncover? Here the news is more reassuring at the global level and more troubling at the level of individual countries. When the authors pooled effects meta-analytically across all 75 systems, three motivational constructs drawn from expectancy-value theory, students’ self-perceived reading competence, their enjoyment of reading, and their attitude toward learning, were consistently and positively associated with resilience under every operationalization and threshold, and girls consistently outperformed boys. Read that result alone, and the choice of definition seems inconsequential.

The country-by-country picture tells a different story. When each education system was analyzed separately, as is standard in PISA-based research, findings were consistent across all nine binary specifications in only 20 percent of systems for gender, 40 percent for attitude toward learning, 69 percent for reading enjoyment, and 72 percent for self-perceived competence. Just two of the 75 systems produced fully consistent results for all four predictors. Threshold level alone flipped conclusions in somewhere between 3 and 37 percent of countries depending on the factor. Much of this inconsistency reflects statistical power, since stricter thresholds shrink samples and smaller-effect predictors, notably gender and attitude toward learning, are the least stable. But the practical consequence does not depend on the cause: a researcher studying one country under one definition could legitimately conclude that a factor matters there when a defensible alternative would say it does not.

The most conceptually striking result came from a complementary interaction analysis that requires no thresholds at all. By modeling whether the socioeconomic gradient in achievement flattens at higher levels of each factor, the authors tested what the term protective factor actually implies. The findings complicate comfortable assumptions: the gradient was flatter, not steeper, among students with higher self-perceived reading competence and stronger attitudes toward learning, but it was significantly steeper, not flatter, among students who enjoyed reading more. In other words, reading enjoyment, though positively associated with achievement on average, was linked to wider rather than narrower socioeconomic gaps, a pattern that no threshold-based definition could ever detect. These moderation effects also varied in direction across countries, appearing in only a minority of systems individually.

The authors close with four practical recommendations: match the operationalization to the research question, report sensitivity analyses at a minimum of two thresholds, present cross-country rankings under multiple definitions, and document or pre-register operationalization decisions so findings can be interpreted in light of the choices that produced them. They caution that their analysis is cross-sectional and cannot establish causation, that their predictors were individual-level only, and that PISA’s socioeconomic index has itself attracted methodological criticism. Still, the broader message is hard to escape. Academic resilience, as measured in large-scale assessments, is not a fixed quantity waiting to be counted but a construct whose observed properties depend partly on how it is defined, and both researchers and policymakers ignore that dependence at their peril.

Subject of Research: How operationalization and threshold choices affect estimates of academic resilience and its protective factors across 75 education systems using PISA 2018 data.

Article Title: Academic resilience in 75 education systems: how operationalization and threshold choices shape findings on protective factors

Article References: Žáková, M., & Lintner, T. (2026). Academic resilience in 75 education systems: how operationalization and threshold choices shape findings on protective factors. Large-scale Assessments in Education, 14(1), Article 44. https://doi.org/10.1186/s40536-026-00318-6

Image Credits: AI Generated

DOI: 10.1186/s40536-026-00318-6

Keywords: academic resilience, PISA 2018, socioeconomic status, educational equity, large-scale assessment, protective factors, operationalization, thresholds, student motivation, expectancy-value theory, educational measurement, cross-national comparison