Screen time myths: what the large studies actually find

The problem with big survey datasets is that they contain hundreds of variables, and there are thousands of defensible ways to analyse them. Choose one path and you can find harm; choose another and you can find nothing. Orben and Przybylski tackled this with a technique called specification curve analysis, which runs every reasonable analysis and reports the whole range. Across three large datasets covering 355,358 adolescents, the association between digital technology use and well-being was negative but small, explaining at most 0.4% of the variation in well-being.

To give that number a sense of scale, the authors compared it with other factors in the same data. Being bullied and smoking cannabis showed far stronger links with poorer well-being, while everyday factors such as regularly eating potatoes or wearing glasses showed links in the same small range as technology use. Their conclusion was that effects this small were too slight to justify policy change on their own.

It helps to picture what 0.4% of the variation means. Imagine lining up every teenager in the data by well-being. Knowing how much each one used technology would let you predict their place in the line only very slightly better than guessing. Family, friendships, sleep, health and school would all tell you far more. That is not the same as saying technology use never matters for anyone, but it is a long way from the claim that screens are the main force shaping a generation’s mental health.

Sources1Source 1Orben, A., & Przybylski, A. K. (2019a). The association between adolescent well-being and digital technology use. Nature Human Behaviour, 3(2), 173–182.

A preregistered study of 120,115 English adolescents by Przybylski and Weinstein tested what they called the Goldilocks hypothesis: that too little and too much use might both be worse than a moderate amount. The relationship between screen time and mental well-being was curved rather than a straight downward line, and it differed between weekdays and weekends. Moderate use was not associated with harm and may be an advantage in a connected world.

That fits a common-sense view of screens as part of ordinary life. A teenager who never goes online is missing out on something too, including a good deal of the social life of their friends.

It also reframes the most important practical risk. Where screens do seem to cost something, the likeliest cost is what they push aside, especially sleep. Late-night use that eats into rest is worth addressing whatever it does to mood directly, and we cover the evidence in gaming before bed.

Sources2Source 2Przybylski, A. K., & Weinstein, N. (2017). A large-scale test of the Goldilocks hypothesis: Quantifying the relations between digital-screen use and the mental well-being of adolescents. Psychological Science, 28(2), 204–215.

Most studies ask people to estimate their own screen time. Parry and colleagues gathered every study they could find that measured media use both ways, by asking and by logging it on the device, and combined 106 effect sizes in a preregistered meta-analysis. Self-reports correlated only moderately with logs and were rarely accurate. Questionnaires about “problematic” use lined up with the logs even less.

That matters because a study built on inaccurate estimates can find associations that belong to the estimate rather than the behaviour. People who feel low may simply believe they spend longer on their phones. When Orben and Przybylski used detailed time-use diaries instead, in 17,247 young people across Ireland, the United States and the United Kingdom, they found little evidence of substantial negative associations with well-being, including for screen use in the run-up to bedtime.

Sources3Source 3Parry, D. A., Davidson, B. I., Sewall, C. J. R., Fisher, J. T., Mieczkowski, H., & Quintana, D. S. (2021). A systematic review and meta-analysis of discrepancies between logged and self-reported digital media use. Nature Human Behaviour, 5(11), 1535–1547.4Source 4Orben, A., & Przybylski, A. K. (2019b). Screens, teens, and psychological well-being: Evidence from three time-use-diary studies. Psychological Science, 30(5), 682–696.

Small average effects are not the same as no effects. Odgers and Jensen reviewed meta-analyses, large preregistered studies and intensive day-by-day studies of adolescents. Most of the research was correlational and produced a mix of small positive, negative and null findings. The most rigorous studies found small associations that could not separate cause from effect and were unlikely to be of clinical significance on average. Averages can still hide individual young people for whom things go badly, which is why the useful research question is increasingly who is affected, and how, rather than whether screens are good or bad.

Timing within development may matter too. Orben and colleagues analysed data from 84,011 people aged 10 to 80 in the UK and found the link between social media use and life satisfaction was most negative in younger adolescents. In longitudinal data they identified windows of greater sensitivity: around ages 11 to 13 for girls and 14 to 15 for boys, with another window at 19 for both. The relationship also ran the other way, with lower life satisfaction predicting more social media use later.

Sources5Source 5Odgers, C. L., & Jensen, M. R. (2020). Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. Journal of Child Psychology and Psychiatry, 61(3), 336–348.6Source 6Orben, A., Przybylski, A. K., Blakemore, S.-J., & Kievit, R. A. (2022). Windows of developmental sensitivity to social media. Nature Communications, 13(1), 1649.

The research consistently points away from a single number and towards context. It also suggests caution about dramatic fixes: as we found when looking at whether a digital detox works, trials of simply switching off have produced mixed results. What is on the screen? What would you be doing otherwise? Is it the hour you would have slept, the conversation you would have had, or the dull wait in a queue? How do you feel afterwards, and do you feel able to stop?

Games follow the same pattern. Large studies using logged play data have found that hours played says little about players’ well-being, while how play fits into their lives says more; we cover that in is gaming good or bad for your mood?. It is also why we write about designing for the life players actually have rather than chasing more minutes of play. If you want to know how we choose and read studies like these, our note on how the OutBrick Journal reads research explains the rules we follow.

• Be wary of any claim built on a single number of hours.
• Look at what the time replaces, especially sleep, movement and time with people.
• Check the device’s own screen-time report rather than your estimate.
• Pay particular attention in early adolescence, when sensitivity may be higher.

Sources3Source 3Parry, D. A., Davidson, B. I., Sewall, C. J. R., Fisher, J. T., Mieczkowski, H., & Quintana, D. S. (2021). A systematic review and meta-analysis of discrepancies between logged and self-reported digital media use. Nature Human Behaviour, 5(11), 1535–1547.5Source 5Odgers, C. L., & Jensen, M. R. (2020). Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. Journal of Child Psychology and Psychiatry, 61(3), 336–348.6Source 6Orben, A., Przybylski, A. K., Blakemore, S.-J., & Kievit, R. A. (2022). Windows of developmental sensitivity to social media. Nature Communications, 13(1), 1649.