Screen Time and Bullying: No Study Has Settled Which Comes First
What 117 longitudinal studies establish about screens and children's socioemotional problems, why the cyberbullying data cannot order the events, and the one activity where the effect is far larger.
By Gale Staff · November 21, 2026 · Psychological Bulletin
Not reviewed by a clinician.

The short answer
No study located here tracks bullying and screen time in the same children over time, so for bullying specifically the question of which comes first is unsettled. The longitudinal evidence that does exist concerns socioemotional problems more broadly: a meta-analysis of 117 studies covering 292,739 children found small associations running in both directions, a coefficient of 0.06 each way, with the figures for gaming several times larger in both directions. The caveat is that the strongest cyberbullying data, an analysis of 9,686 early adolescents, was measured at a single point in time, which means it cannot establish what came before what.
Who to see
Pediatrician
No study has established whether screens come before bullying or follow it, and the effects either way are small; a pediatrician can talk through your child's screen habits.
Find care →The question a parent is actually asking
A boy of twelve has been quiet since half term. His mother knows two things and cannot work out the order they belong in. He is being left out of something at school, and he is on his tablet far more than he was in the summer. The headline she reads that week tells her that bullying is linked to more screen time, which she already knew, and implies a direction, which is the only part she needs.
The direction matters because the two readings call for different things. A child retreating to a screen because the week has been unbearable is a child carrying a social problem. A child whose hours online are feeding what happens to him is a different situation. The evidence has something careful to say about this, and it is not the thing the headline says.
What the strongest study measured
The most serious attempt to put screens and children's wellbeing in time order is a systematic review and meta-analysis by Vasconcellos and colleagues, published in Psychological Bulletin in 2025. Its premise is the problem itself. Most research in this field, the authors note, has been cross-sectional, which they say makes it difficult to establish causality: if you measure a child's screen use and a child's distress on the same afternoon, you learn that they travel together and nothing about which arrived first.
So they went looking only for studies that measured children more than once. A total of 132 longitudinal studies met the inclusion criteria and entered the systematic review, and from those, 117 studies were meta-analysed, covering 292,739 children and 2,284 separate effects. That pair of numbers is sometimes reported as a contradiction in coverage of the paper. It is not one: the larger number is the review and the smaller one is the subset that could be pooled statistically.
Both arrows exist, and both are small
The finding is that both directions are real and neither is large. Screen use led to socioemotional problems with a coefficient of 0.06, confidence interval 0.02 to 0.11. Socioemotional problems led to greater screen use with the same coefficient of 0.06, confidence interval 0.01 to 0.12, across 200,018 children. In other words the mother in the kitchen is right twice over, and the size of what she is right about is modest in both directions.
A word of caution about that 0.06. It is a standardised coefficient, and the paper's abstract does not translate it into anything a parent could picture, such as hours or symptoms. It is best read as establishing that the relationship exists and runs both ways, rather than as a measure of how much.
The exception is the interesting part. For gaming specifically, the coefficients were several times larger: socioemotional problems led to more gaming at 0.44, confidence interval 0.29 to 0.60, and playing games led to later socioemotional problems at 0.32, confidence interval 0.23 to 0.42, both across 80,809 children. The authors' own conclusion is that their temporal evidence reinforces the benefits of screen time guidelines but suggests a change in focus, which is a quiet way of saying that what a child is doing on the screen may matter more than the total on the clock.
The bullying evidence cannot put events in order
None of the above measured bullying. The closest large study is an analysis by Nagata and colleagues in the Journal of Adolescent Health in 2026, drawing on 9,686 early adolescents aged 11 to 15 in the Adolescent Brain Cognitive Development study, and its own first sentence about method is the limitation: it analysed cross-sectional data, measured at one point.
What it found is a set of associations with how screens are handled at home. Parental allowance of screen use at mealtimes carried adjusted odds of cyberbullying victimisation of 1.14, screen use in the bedroom 1.31, and screen use as a way of managing behaviour 1.18. Allowance of bedroom screen use was also associated with higher odds of being the one doing the cyberbullying, at 1.44. Greater restriction of screen time went the other way, with odds of victimisation of 0.80. Daily active screen time partially mediated these associations, and the associations were stronger for being bullied than for bullying others.
Because every one of those numbers comes from a single moment of measurement, none of them can say whether the bedroom tablet preceded the trouble or followed it. A scoping review of 24 studies published in Behavioral Sciences in 2026 reaches for the same point when it calls for further longitudinal research to clarify the temporal relationships around cyberbullying victimisation.
Where a smaller study points
One cohort has looked at the ordering directly, though not at bullying. Soto-Ramirez and colleagues followed 669 Chilean children measured twice over two years and used a cross-lagged panel model, the design built for exactly this question. Higher screen time at the first wave did not significantly predict externalising problems at the second, with a coefficient of minus 0.01 and a p-value of 0.71. Externalising problems at the first wave did significantly predict higher screen time at the second, coefficient 0.08, p-value 0.043.
In that cohort, then, only one of the two arrows the big meta-analysis reports actually appeared, and it was the one pointing from the child's difficulties toward the screen. A single two-wave study of 669 children does not overturn 117 studies of 292,739, and it is not offered here as doing so. It is offered because it is the shape of evidence that can answer a question about order at all, and because the honest state of this field is that very few studies have been built that way.
What this study can't tell you
- Whether screens cause bullying or bullying drives screen use. No study located here follows bullying and screen time in the same children over time; the longitudinal evidence concerns socioemotional problems in general, and the cyberbullying evidence was measured at a single point.
- What a coefficient of 0.06 means in a child's week. The meta-analysis reports it in each direction without translating the scale in its abstract, so the figure establishes direction and statistical significance rather than a magnitude anyone can picture.
- Whether social media in particular is the driver. The cyberbullying analysis measured parental media practices and daily active screen time; social media was not reported as a separate exposure in its abstract.
- Whether restricting screens reduces bullying. The link between greater restriction and lower odds of victimisation, odds of 0.80, was measured at one point in time, so it cannot show that changing the rule changes the outcome.
- Why one cohort found only half the pattern. In 669 Chilean children, screen time did not predict later externalising problems while externalising problems did predict later screen time, and the reason for the difference from the pooled result is not established by either paper.
- What a decade of early-life screen use does to bullying. A long-term follow-up was described in the coverage behind this page, but it could not be located in the literature index, so nothing here rests on it.
The Gale read
The headline that sends a parent looking gets the arrow wrong by implying there is one. On the best temporal evidence available, screens and children's socioemotional difficulties feed each other in both directions, and both effects are small enough that neither can carry the weight a worried evening puts on it. For bullying specifically the position is plainer and less satisfying: the large study that can order events in time did not measure bullying, and the large study that measured bullying could not order events in time. What survives that is worth having anyway. The associations cluster around where and how a screen is used rather than the hours on it, with the bedroom showing up in both victimisation and perpetration, and gaming is the one activity where the numbers stop being small in either direction. The field's own reviewers keep asking for longitudinal work on exactly the question parents ask first, which is the clearest signal available that it has not yet been answered. A parent reading a figure about bullying and screen time this week is reading a real association and an unknown sequence, and treating the second as settled is where the harm in this topic currently lies.
Common questions
Does screen time cause bullying?
The evidence cannot say. No study located here measures bullying and screen time in the same children over time, which is what establishing a cause would require. The nearest large longitudinal evidence concerns socioemotional problems rather than bullying, and it finds small effects running in both directions, a coefficient of 0.06 each way across 117 studies and 292,739 children.
Is my bullied child using screens more?
That direction has support, and it is the half of the relationship often left out. In the pooled longitudinal evidence, socioemotional problems led to greater screen use at a coefficient of 0.06 across 200,018 children, and for gaming the figure was 0.44. In one two-wave cohort of 669 children, externalising problems predicted later screen time while screen time did not predict later externalising problems. None of those studies measured bullying itself.
What is the link between screen time and cyberbullying?
The largest analysis, covering 9,686 adolescents aged 11 to 15, reports associations with how screens are handled rather than with hours alone: adjusted odds of being cyberbullied were 1.31 where bedroom screen use was allowed, 1.14 at mealtimes and 1.18 where screens were used to manage behaviour, while greater restriction carried odds of 0.80. Bedroom screen use was also associated with doing the cyberbullying, at 1.44. All of it was measured at a single point, so none of it orders the events.
Does social media cause bullying in kids?
No study located here isolates social media as its own exposure. The cyberbullying analysis measured parental media practices and daily active screen time, and the longitudinal meta-analysis separated out gaming, where the effects were largest in both directions, rather than social media. So the specific claim about social media is not what this evidence tested.
Sources
- 1.Vasconcellos RP, Sanders T, Lonsdale C, Parker P, Conigrave J, Tang S, Del Pozo Cruz B, Biddle SJH, Taylor R, Innes-Hughes C, Salmela-Aro K, Vasconcellos D, Wilhite K, Tremaine E, Booker B, Noetel M. Electronic screen use and children's socioemotional problems: A systematic review and meta-analysis of longitudinal studies. Psychological Bulletin. 2025;151(5):513-543. doi:10.1037/bul0000468 link
- 2.Nagata JM, Yen F, Paul A, Kim KE, Shao IY, Al-Shoaibi AA, Ganson KT, Testa A, Kiss O, He J, Groves AK, Baird S, Baker FC. Media Parenting Practices and Early Adolescent Cyberbullying: Findings From the Adolescent Brain Cognitive Development Study. Journal of Adolescent Health. 2026;79(3):382-389. doi:10.1016/j.jadohealth.2026.04.021 link
- 3.Soto-Ramirez P, Godoy F, Narea M, Ayala C. Screen exposure in Chilean children during early childhood and socio-emotional problems: relationship and directionality. Frontiers in Psychology. 2025;16:1589113. doi:10.3389/fpsyg.2025.1589113 link
- 4.Navarro-Vitoria A, Patro-Hernandez RM, Galvez-Sanchez CM, Liminana-Gras RM. Factors and Psychosocial and Physical Health Outcomes Associated with Cyberbullying Victimization in Adolescents: A Systematic Scoping Review. Behavioral Sciences (Basel). 2026;16(10):1733. doi:10.3390/bs16101733 link
4 sources, numbered by first appearance. General health information, not medical advice. AI-assisted editorial content — every citation independently verified. Editorial policy
More from Gale Staff
Research
Top-Third Vitamin D Came With 19 More Minutes of Sleep a NightResearch
A Stroke Drug Beat Placebo, 69.7% to 56.3%; An Earlier Trial Did Not
Published November 21, 2026 · Gale Staff