A study conducted among Lombardy’s high school students has found that those who created a personal social media account in 6th, 7th or 8th grade tended to score lower on standardised tests in Italian and mathematics than peers who waited until 9th grade or later. The earlier the start, the larger the observed gap, according to researchers behind the paper published in Nature Human Behaviour.
The question of when children should be allowed to begin using smartphones and social media has become a heated topic among scientists, educators and parents. Many pupils acquire their first smartphones at around 10 or 11, and frequently establish social media accounts before reaching platform or national minimum ages. This trend has raised concerns, given growing links between digital media use and indicators of mental health and academic performance in several countries.
Researchers led by Marco Gui of the University of Milano-Bicocca examined whether early social media use harms school performance by distracting pupils, displacing time for homework, sleep or physical activity, and by exposing them to content and relationships that may affect well-being. They also anticipated stronger negative effects with longer exposure.
The study combined two data sources: the EYES UP survey, a retrospective questionnaire completed by 6,609 students in 10th and 11th grade in Lombardy during the 2023–2024 school year, and data from INVALSI, the Italian National Institute for the Evaluation of the Education and Training System, which administers standardised tests in Italian and mathematics across multiple grades.
Following the merge of datasets and focusing on students born in 2007 and 2008, the analytic sample amounted to 5,227 pupils. The survey assessed when pupils opened their first social media profile and first internet-connected smartphone, as well as other technology use, daily life intrusion of smartphones, well-being, life satisfaction and academic self-efficacy.
The results show varying patterns in device adoption. By the end of 8th grade, almost all pupils—98%—owned a smartphone, with many beginning to use messaging apps around the time they received their device or the following year. In terms of social media, 11% opened their first profile in primary school, 30% in 6th grade, 25% in 7th grade and 18% in 8th grade, while only 16% waited until 9th grade, the minimum age for independent access under Italian law.
Students who started using social media earlier tended to have lower scores in Italian and mathematics, even after adjusting for other technology exposure. Early adopters were also more likely to come from less-educated families, less likely to live with both parents, and to have lower test scores prior to opening a social media account.
To better isolate the role of social media, the researchers paired pupils who started in 6th, 7th or 8th grade with those who waited until 9th grade or later, matching on family background, prior technology use and 5th-grade performance, then tracked changes over time using a difference-in-differences approach. The strategy aimed to compare similarly positioned pupils who differed only in the timing of their social media use.
Among these comparisons, pupils who opened an account in 6th grade scored about 0.22 standard deviations lower in Italian and 0.18 lower in mathematics by 8th grade than their matched peers who waited. By 10th grade, the Italian gap remained at about 0.22 standard deviations, while the mathematics gap widened to roughly 0.27 standard deviations. For those who started in 7th grade, the gaps were in the same direction but smaller; for 8th grade starters, gaps were even smaller, with the mathematics difference not statistically significant.
The researchers note that the findings for 8th-grade starters were less robust in supplementary analyses, while 6th-grade starters produced the most robust results. Notably, early social media use did not correlate with English test scores, a finding the authors suggest may reflect the high volume of English-language content online. Early adopters were also more likely to repeat a grade and reported lower academic self-efficacy.
An exploratory analysis suggested that the extent to which smartphones permeate daily life—checking devices before bed, on waking, or during homework—accounts for roughly 15% to 47% of the link between early social media use and lower test scores, depending on the subject. The authors emphasise that this analysis is correlational in nature.
“The findings of this study carry important implications for educators, parents and policymakers. First, they emphasise the need for more structured guidelines and digital literacy programmes in schools aimed at developing appropriate awareness of social media use among young students and parents, particularly those from disadvantaged backgrounds. Second, they suggest the need to regulate the engagement mechanisms embedded in these platforms, particularly when they are used by minors. Moreover, in the short term, policies that delay social media access until children are cognitively mature enough to manage its potential distractions could be beneficial. Some of these policies are already in place,” the study authors concluded.
The study adds to the body of evidence on social media use and academic outcomes, yet the authors caution that the size of the observed differences can be interpreted in different ways. They note that effects above 0.2 standard deviations are regarded as large in education research and roughly equivalent to six months of schooling, though in psychology such a gap is typically considered small.
Limitations include that the study was not a randomised experiment, so unmeasured factors—such as family dynamics—that influence both social media use and academic performance could affect results. Age of first use was based on pupil recall, introducing potential recall bias, and the sample excluded pupils in short vocational programmes or those who had dropped out or repeated multiple years.
The authors also acknowledge that the findings may not generalise beyond northern Italy or to the social media landscape of 2019 to 2023, when the participants were adolescents. Two of the authors, Gui and Chiara Respi, sit on the governance board of the Patti Digitali Foundation, which advocates a gradual approach to children’s access to digital environments.
The paper, “Longitudinal effect of early social media use on standardized learning outcomes during school career,” lists Marco Gui, Chiara Respi, Giovanni Abbiati, Cristiana Paladini, Sofia Ercolanoni, Francesca Milzani, Tiziana Pirola and Giovanni Vezzoli as authors.
