A multivariate twin model of screen time at age 11 and social network size at ages 11 and 15.

Two important and interlinked areas during childhood and adolescence are screen media use and peer relationships. A wide range of evidence and research has explored the effects of screen and media use in children and adolescents, finding significant associations between screen time use and detrimental social outcomes across development, as well as a wide range of cognitive and emotional outcomes. However, little research has been conducted on the relative influences of genetic and environmental factors on screen time, although that which has engaged with the topic has found substantial heritability for individual differences in screen use (Ayorech et al. 2017). Additionally, although some research has examined genetic effects on peer relationships (Boivin et al., 2013), none has explicitly focused on genetic and environmental influences on individual differences in social network size specifically. This genetically informed study therefore will fill and clarify important gaps in the literature by examining the relative contributions of genetic and shared and nonshared environments on both screen time and size of social network. There is strong evidence of a link between screen media use and poorer psychosocial outcomes, such as early academic or cognitive skills, and social or behavioural outcomes (Carson et al. 2015; Stiglic & Viner 2019). While much research in this area has tended to focus on younger children, even examining those at pre-school age (e.g. Aishworiya et al. 2022), less research explores effects during adolescence, which this project will focus on. There is substantial evidence suggesting screen media use is associated with peer relationships and social skills (e.g. Gath et al., 2025). Generally, this research finds negative associations, for example Yu et al. (2025) found that excessive screen time was significantly associated with peer relationship issues. However, some studies have found that increased screen time, particularly when social media is considered, is actually associated with having a larger social network. These moderately mixed findings reinforce why further research is needed in this area. The present study will fill this gap by examining phenotypic and multivariate twin associations between screen time and size of social network. As a result of the mixed findings, our hypothesis about the direction of association is somewhat exploratory, although on balance appears more likely to be a negative association, particularly as we are examining screen/internet time more broadly, rather than phone/social media use in particular. More broadly, heritability studies estimate that around 50% of variance in many traits is attributable to genetic factors. Although many behavioural traits show moderate heritability, the heritability of individual traits cannot alone imply that associations between these traits are genetically mediated. However, evidence from genetically informed studies indicates that links between screen time and related behavioural outcomes, such as psychiatric and socio-behavioural traits, may be partly explained by shared genetic influences (e.g. Zhang et al. 2023). Therefore, it seems likely that associations between screen media use and social network size may likewise reflect shared genetic influences, a hypothesis that this project is explicitly designed to test. This project uses data from the TwinLife dataset conducted in Germany, a longitudinal twin family study on the development of social inequality. This data contains four twin birth cohorts, each with around 1000 same-sex twin pairs, who are studied in both face-to-face and telephone interviews. This allows a behavioural genetics approach to be taken, using bivariate genetic analysis to study the extent to which the links between screen time use and social network are drive by genetics, the shared environment and the nonshared environment.