Social media has become deeply woven into adolescent life worldwide. Global surveys cited in the study, including the Digital 2024 overview report and Pew Research Center data on teens and technology, document near-universal engagement with social platforms among young people, while the World Health Organization’s Health Behaviour in School-aged Children survey has raised concerns about adolescent mental health and well-being in the digital era. At the same time, a growing body of research links problematic social media use to depression, anxiety, loneliness and disrupted identity development. Understanding which family dynamics feed into these addiction-like patterns, and for whom, has become a pressing question for clinicians and educators.

Previous studies have consistently connected family functioning to problematic internet and social media use. Work on attachment styles and family climate has shown that adolescents from less cohesive households are more likely to develop compulsive digital habits, and longitudinal research among Chinese university students found that poor family functioning predicted later problematic social media use, partly through loneliness. A recent systematic review concluded that family conflict is associated with an increased risk of problematic digital media use in children and adolescents. But most of this literature treats family functioning as a single global variable, and few studies have examined how its individual components interact with specific addiction symptoms at the level of the individual adolescent.

The new study took a different analytical route. Instead of asking whether family functioning predicts addiction on average, Aldemir applied psychometric network analysis, a framework rooted in network theories of mental disorders. In this approach, psychological variables are modeled as a web of mutually reinforcing nodes rather than as effects of a single latent cause. The researcher used a Gaussian Graphical Model, estimated with regularized partial correlations, to map the statistical relationships among sub-dimensions of the McMaster Family Assessment Device, a widely used measure of family health, and symptoms of social media addiction assessed with the Bergen Social Media Addiction Scale.

The sample consisted of 966 adolescents in Türkiye, 537 girls and 429 boys, aged 13 to 19. To ensure that the estimated network was not an artifact of sampling noise, the study employed bootstrapping with 1,000 replicates to test the stability and accuracy of the network’s edges and centrality indices. The stability coefficient reported for the network was described as excellent, meaning the key findings were unlikely to shift dramatically under resampling. Differences between the boys’ and girls’ networks were formally tested using the Network Comparison Test, which evaluates whether the structure and connectivity of two estimated networks differ significantly.

Across the whole sample, one family dimension stood out. Affective involvement, the degree to which family members are emotionally connected and responsive to one another, emerged as the most central node in the global network, suggesting it occupies a hub position from which dysfunction can propagate to other family processes. However, centrality within the family system was not the same as being the doorway to addiction. When the researcher computed bridge centrality, a metric designed to identify variables that connect otherwise separate communities of symptoms, problem solving proved to be the most critical bridge. In other words, a household’s inability to work through conflicts and practical difficulties appeared to be the primary channel through which family dysfunction translated into addiction symptoms.

The gender comparison revealed a striking asymmetry. For girls, the pathway linking family environment to social media addiction ran most clearly through problem-solving deficits and difficulties in emotional processing. For boys, the strongest connection involved deficits in behavioral control, the family’s capacity to set rules, monitor conduct and maintain consistent standards. This pattern aligns with broader findings in the addiction literature, including meta-analytic evidence that boys and girls tend to gravitate toward different flavors of problematic internet use, with gaming and control-related vulnerabilities more prominent among males and relational and emotion-focused patterns more prominent among females. It also echoes longitudinal work showing that maladaptive cognitions explain gender differences in problematic gaming and social media use.

The theoretical implications reach beyond statistics. Family systems theory, from Bowen’s work on differentiation and emotional triangles to the McMaster model operationalized in the Family Assessment Device, has long held that families function as integrated emotional units. Bronfenbrenner’s ecological model similarly places the family at the center of adolescent development. The new findings add a network-level nuance to these traditions: the same family system can channel risk into different behavioral outcomes depending on the adolescent’s gender, and the specific symptom-level connections involved may differ accordingly. From a self-determination perspective, dysfunctional families may fail to satisfy adolescents’ needs for autonomy, competence and relatedness, pushing them toward digital environments that promise both escape and connection, but the route by which that push occurs appears gendered.

The practical consequences are potentially significant. If problem-solving deficits are the main bridge for girls, then family-based interventions that coach collaborative conflict resolution and emotional expression may be the most effective lever for reducing social media addiction risk in daughters. If behavioral control deficits are the main bridge for boys, then interventions emphasizing structure, consistent rule-setting and parental monitoring may matter more for sons. Recent meta-analytic work on parental control and problematic internet use supports the general importance of the monitoring dimension, and network studies of parental monitoring across developmental stages similarly suggest that specific parenting behaviors, rather than global family sentiment, may drive risk. The study’s author concludes that clinical and preventive efforts should be tailored specifically: emotion- and solution-oriented for girls, and structure- and control-oriented for boys.

As with any cross-sectional network study, caution is warranted in drawing causal conclusions. The data capture associations among self-reported symptoms at a single point in time, and network edges do not by themselves prove directional influence. The sample was drawn from Türkiye, and cultural context may shape both family dynamics and patterns of digital use, although Turkish survey data from the national statistical institute indicate high household internet penetration comparable to other middle- and high-income settings. Still, the methodological rigor, the large sample, the bootstrapped stability checks and the formal network comparison give the findings unusual weight for this literature. The dataset has been made publicly available on Harvard Dataverse to support transparency and replication. If the gender-specific pathways hold up in longitudinal and cross-cultural research, the study could mark a shift in how clinicians think about digital addiction prevention: away from one-size-fits-all family programs and toward interventions matched to the distinct psychological routes by which family distress becomes compulsive scrolling.

Subject of Research: Gender-specific pathways linking family functioning sub-dimensions to social media addiction symptoms in adolescents

Article Title: Problem solving for girls, behavioral control for boys: gender-specific pathways linking family functioning to social media addiction

Article References: Aldemir, A. (2026). Problem solving for girls, behavioral control for boys: gender-specific pathways linking family functioning to social media addiction. Current Psychology, 45(17), Article 1451. https://doi.org/10.1007/s12144-026-10012-z

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10012-z

Keywords: social media addiction, family functioning, adolescence, network analysis, gender differences, behavioral control, problem solving, psychometric network analysis, Bergen Social Media Addiction Scale, family assessment, parental monitoring, Türkiye