The study recruited adolescents aged 13 to 18 who were attending high schools in a provincial center in eastern Türkiye. Participants completed a demographic information form along with three validated instruments: the Bergen Social Media Addiction Scale, which measures the core features of behavioral addiction such as salience, tolerance, mood modification, and conflict; the Friendship Qualities Scale, which assesses dimensions of peer relationships such as companionship, help, intimacy, and conflict; and the Adolescents’ Attitudes Toward Violence Scale, which captures how acceptable young people consider violent behavior to be. Ethical approval was obtained from the Ardahan University Non-Interventional Research Ethics Committee, with written informed consent from parents or legal guardians and assent from the adolescents themselves, and the study was conducted in accordance with the Declaration of Helsinki.

The headline numbers are striking in their ordinariness. The mean Bergen Social Media Addiction Scale score in the sample was 17.18, with a standard deviation of 5.30 — a figure suggesting that, on average, these adolescents were not in the extreme range of problematic use, though individual scores varied considerably. The mean peer relations score was 82.15, and the mean attitude toward violence score was 24.69. What matters for the study’s central question is not the raw averages but the correlations between them, and it is there that the researchers found their most interesting result.

Using Pearson correlation analysis, the team found a positive and statistically significant association between social media addiction scores and attitudes toward violence: the correlation coefficient was 0.214, with a p-value below .001. In practical terms, adolescents who reported more compulsive, dependency-like engagement with social media platforms were also more likely to view violence as acceptable. The correlation is modest — a coefficient of 0.214 means the two variables share only a small fraction of their variance — but in a sample of 450, it is unlikely to be a statistical fluke. Importantly, a cross-sectional design cannot establish which way the arrow points, or whether a third factor drives both. Adolescents already prone to aggression may gravitate toward certain online content, or the content itself may shape attitudes, or both processes may reinforce each other over time.

The second major finding was, in a sense, the absence of one. The correlation between social media addiction and the quality of peer relationships was essentially zero, at r = 0.004, which was not statistically significant. This null result challenges a common intuition: that teenagers who are glued to their phones must be sacrificing the depth and warmth of their real-world friendships. In this sample, at least, problematic social media use and friendship quality appeared to operate on independent tracks. A teenager could score high on social media addiction while still reporting strong, supportive friendships — or low on addiction with strained peer relations — without any systematic link between the two.

To test these relationships more rigorously, the researchers turned to multiple linear regression, a statistical technique that estimates the unique contribution of each predictor while holding the others constant. When both social media addiction and peer relationship quality were entered as predictors of attitudes toward violence, only social media addiction remained a significant factor, with a standardized beta coefficient of 0.126 and a p-value below .001. Peer relations, once the shared variance was accounted for, showed no significant association with violent attitudes. The picture that emerges is one in which problematic social media use stands apart as the variable of interest, while friendship quality — however important it may be for other aspects of adolescent development — did not predict attitudes toward violence in this model.

The researchers were candid about the limits of their explanatory power. The regression model explained just 4.2 percent of the variance in the outcome, an R-squared of 0.042, which the authors themselves describe as limited. This is an important caveat for anyone tempted to over-interpret the findings. Attitudes toward violence in adolescence are shaped by a vast web of influences — family environment, socioeconomic conditions, exposure to violence at home or in the community, personality traits, gender norms, and media consumption far beyond social platforms. A single model built on two psychological variables was never going to account for much of that complexity. What the study offers is not a comprehensive explanation but a targeted, statistically supported association that justifies deeper investigation.

Why might social media addiction and violent attitudes travel together? The authors suggest that problematic social media use should be understood not merely as a social phenomenon but in relation to behavioral and psychosocial factors. Several mechanisms are plausible and consistent with the broader literature. Compulsive use often involves prolonged exposure to algorithmically curated content, which can include aggressive, degrading, or violent material. The disinhibiting anonymity of online spaces may normalize hostile communication styles that bleed into offline attitudes. Sleep disruption, mood dysregulation, and impulsivity — all documented correlates of problematic digital use — are themselves linked to irritability and aggressive responding. And for some adolescents, social media may become a stage for relational aggression, cyberbullying, and status contests in which hostility is rewarded with attention. None of these pathways was tested directly in this study, but each offers a testable hypothesis for future longitudinal work.

The null finding on peer relationships deserves its own reflection. Friendship quality is often treated as a protective factor against a wide range of adolescent risk behaviors, and it may well be. But this study suggests that protective friendships do not necessarily buffer the association between compulsive social media use and violent attitudes — or, put differently, that the mechanism linking heavy social media use to violent attitudes does not run through the quality of adolescents’ offline friendships. That distinction matters for intervention design. A program that strengthens peer bonds may improve wellbeing in many ways, yet the data here hint that it may not, by itself, blunt the attitudinal correlates of problematic digital habits.

The practical implications the authors draw are correspondingly focused. They point to school-based interventions and digital literacy programs as promising avenues for promoting healthier social media use among adolescents. Digital literacy, in this framing, is not just about spotting misinformation; it encompasses the self-regulatory skills needed to recognize compulsive patterns of use, critically evaluate the content a feed delivers, and understand how online experiences can shape emotions and attitudes. Because the study was conducted in a single provincial center with a cross-sectional design, generalization requires caution — attitudes toward violence and patterns of platform use vary across cultures, regions, and age bands. Still, the study’s methodological transparency, its use of validated instruments, and its honest reporting of a small effect size make it a useful building block. As researchers worldwide grapple with the mental and behavioral consequences of an always-online adolescence, findings like these — modest, carefully bounded, but statistically firm — are exactly the kind of evidence that accumulates into understanding. The next step is longitudinal and experimental work that can disentangle whether compulsive scrolling nudges attitudes toward violence, whether pre-existing attitudes drive the scrolling, or whether both spring from a common psychological root.

Subject of Research: The association between social media addiction, peer relationship quality, and attitudes toward violence among adolescents

Article Title: Associations between social media addiction, peer relationships and attitudes toward violence among adolescents: a cross-sectional study

Article References: Dursun, E., Oluç, T., Arga, Z., Akyol, T., & Çiftci, K. (2026). Associations between social media addiction, peer relationships and attitudes toward violence among adolescents: a cross-sectional study. BMC Psychology. https://doi.org/10.1186/s40359-026-05699-5

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05699-5

Keywords: adolescents, social media addiction, peer relationships, attitudes toward violence, cross-sectional study, Bergen Social Media Addiction Scale, digital literacy, adolescent psychology, problematic social media use, regression analysis, Türkiye, school-based interventions