Emotional Skills Online Shield Teens From Cyberbullying, Study Finds

Cyberbullying has become one of the most pressing dangers of adolescence, and a new study from Spain suggests that the key to preventing it may lie not in monitoring screens or banning apps, but in a set of skills that most schools have never taught: the ability to recognise, regulate, and control emotions specifically in digital environments. Researchers at the University of Zaragoza found that certain online-specific emotional competencies strongly predict whether a teenager will attack or be attacked on social media, and that these protective effects differ sharply between boys and girls and between younger and older students.

The study, published in the Journal of New Approaches in Educational Research, surveyed 945 Spanish adolescents aged 12 to 16 enrolled in 13 secondary schools across the Autonomous Community of Aragon. The sample was nearly evenly split by sex, with 56.5 percent girls and 43.5 percent boys, and was distributed across all four years of compulsory secondary education, the stage Spanish students must attend by law. Participants completed two validated self-report questionnaires: the European Cyberbullying Intervention Project Questionnaire, which measures cybervictimisation and cyberaggression, and the Socio-Emotional E-Competencies Questionnaire, known as e-COM, which was designed specifically to assess emotional skills as they operate online rather than in face-to-face settings.

The concept of socio-emotional e-competencies is central to the research. Unlike traditional emotional intelligence, which is usually measured in physical social contexts, e-competencies capture how young people manage their feelings amid the distinctive conditions of online communication: anonymity, disinhibition, the absence of non-verbal cues, and a flood of new expressive forms such as memes, stickers, and viral challenges. The e-COM questionnaire measures five such abilities: emotional e-awareness, the capacity to identify one’s own emotions in a virtual context; emotional e-regulation, the ability to generate appropriate responses to online situations; e-self-control of impulsivity, the capacity to inhibit spontaneous reactions to stimuli encountered on the internet; emotional e-independence, the ability to feel emotionally secure in online relationships without depending on likes, followers, or reputation; and social e-competency, the capacity to maintain good virtual relationships and behave pro-socially.

The prevalence figures the researchers recorded are striking. Just over half of all participants, 51.3 percent, reported having been insulted or taunted online, 35 percent said rumours had been spread about them, 31.1 percent reported being excluded from social networks or games, and 22.1 percent confirmed they had been threatened. On the aggressive side, 27.6 percent admitted to insults or taunts, 23.9 percent to ignoring or excluding someone, 14.3 percent to spreading rumours, and 9.7 percent to issuing threats. While most of these behaviours occurred only sporadically, between 8 and 11 percent of participants reported being threatened on a weekly or monthly basis, and 3 to 8 percent admitted to repeated aggression at that frequency.

Sex and age mattered from the outset. Boys scored significantly higher than girls on both cyberaggression and cybervictimisation overall, and both behaviours increased as students advanced through the academic years. Girls in the first two years of secondary school, aged roughly 12 to 14, reported the lowest levels of both victimisation and aggression, while boys’ scores were also marked by greater variance. In the domain of e-competencies, girls scored higher on social e-competency, particularly in the upper two academic years, and an intriguing interaction emerged for emotional e-independence: girls in the younger years actually scored higher than boys, but by the third and fourth years their scores had dropped below the boys’ averages. Meanwhile, emotional e-regulation and e-self-control of impulsivity declined in both sexes as students moved through school.

To test whether these competencies actually predict cyberbullying roles, the team built a structural equation model treating the five e-competencies as independent variables and cyberaggression and cybervictimisation as dependent outcomes. Using Mplus software with robust maximum likelihood estimation, they compared successive models across four groups defined by sex and by age band, the first two versus the last two academic years. The most parsimonious model, in which certain regression weights were constrained to be equal across groups, provided the best fit according to the Akaike information criterion, and the results were dramatic in the youngest groups: the model explained 34.8 percent of the variance in cyberaggression among younger girls, and roughly 20 percent of the variance in cybervictimisation among younger boys and girls alike.

The specific protective skills differed by sex in a way that carries clear practical implications. For girls, emotional e-regulation was the decisive factor, significantly predicting both cybervictimisation and cyberaggression across all academic years, with negative coefficients indicating that better regulation meant less involvement in either role. For boys, e-self-control of impulsivity took centre stage: it was a strong protective factor against cybervictimisation in the younger group, with a standardised coefficient of minus 0.457, and against cyberaggression in both younger and older boys, with coefficients around minus 0.30. Emotional e-independence also protected younger girls against both victimisation and aggression, while emotional e-awareness made a smaller contribution to explaining cyberaggression in younger boys. Notably, social e-competency showed no protective relationship at all; in fact, among younger students of both sexes it was positively associated with victimisation, possibly because highly socially aware adolescents interact more often online, or because empathetic students who defend attacked peers become targets themselves.

The authors interpret these sex differences in light of established patterns in how boys and girls behave online. Boys tend to use cyberaggression more directly, often as revenge, which aligns with the central role of impulsivity control; an adolescent who reposts unverified messages or fires off an angry reply without pausing exemplifies the risk. Girls, by contrast, are more likely to aggress through disinhibition and to share personal information with greater emotional intensity, so the ability to regulate emotions, to imagine how a joke will land before posting it, becomes the critical safeguard. The researchers also note that the correlations they observed between e-competencies and cyberbullying were generally stronger than those reported in earlier studies of offline emotional skills, suggesting that competencies measured in the environment where cyberbullying actually occurs provide a more precise picture of the phenomenon.

Perhaps the most consequential finding concerns timing. The e-competencies explained by far the largest share of cyberbullying variance in the youngest students, and both e-self-control of impulsivity and emotional e-regulation were found to decline steadily from the first to the fourth year of secondary school in both sexes. The authors argue this makes a strong case for embedding these skills into the curriculum from early adolescence, before patterns of aggression and victimisation take hold. They suggest incorporating specific e-competency modules into existing emotional education programmes, using collective intelligence apps that have shown beneficial results, and involving families and teachers in the training. Digital literacy programmes could even begin before adolescence, providing positive mentoring as children first encounter the internet.

The study is not without limitations. Its cross-sectional design means causality could run in the opposite direction, with persistent cyberbullying potentially eroding e-competencies rather than the reverse, and longitudinal studies will be needed to settle the question. The self-report measures are vulnerable to social desirability bias, the sample was quota-based rather than random, and only boys and girls were considered. The authors call for replication in other countries and for the inclusion of further variables such as peer support, social pressure, netiquette, and family supervision. Even so, the core message stands out as genuinely new: cyberbullying prevention has largely borrowed tools designed for face-to-face bullying, but the online world has its own emotional grammar, and teaching adolescents to read it, to value themselves without counting likes, to pause before reacting to a notification, and to name what they feel while scrolling, may be the most direct route to reducing both the attacks and the wounds.

Subject of Research: The role of socio-emotional e-competencies in predicting cyberaggression and cybervictimisation among adolescents

Article Title: Socio-emotional e-competencies, cyberaggression, and cybervictimisation in adolescents: differences according to sex and academic year

Article References: Cebollero-Salinas, A., Hernández, S. O., & Escoriaza, J. C. (2025). Socio-emotional e-competencies, cyberaggression, and cybervictimisation in adolescents: differences according to sex and academic year. Journal of New Approaches in Educational Research, 14(1), Article 5. https://doi.org/10.1007/s44322-024-00025-6

Image Credits: AI Generated

DOI: 10.1007/s44322-024-00025-6

Keywords: cyberbullying, cyberaggression, cybervictimisation, socio-emotional e-competencies, adolescents, emotional regulation, impulsivity, social media, structural equation modeling, gender differences, secondary education, digital literacy