Digital technology explains very little of teen wellbeing

The short answer

If you want one number for how much screens affect a typical teenager's happiness, the most careful attempt to find it came back very small. A 2019 reanalysis of three large surveys found that digital technology use explained, at most, four tenths of one percent of the variation in adolescent wellbeing. That is not a verdict on your own household, only a statement about what population averages can show.

What the evidence says

Amy Orben and Andrew Przybylski published their analysis in Nature Human Behaviour, first online on 14 January 2019, using specification curve analysis, a method that runs an outcome through many defensible analytic choices rather than one preferred model. Applied to three datasets covering 355,358 adolescents, technology use measures were compared against wellbeing measures across a wide range of reasonable ways to define and adjust for both. The associations that emerged were, in the authors' words, negative but small, and the paper concluded that the evidence available did not support the kind of confident regulatory claims sometimes built on single-model studies.

For context

This is not the same claim as 'screens are harmless'. A companion paper by Orben, Dienlin and Przybylski, published in PNAS in May 2019 using panel data, described social media's effect on life satisfaction as small at best, running in both directions over time, and different for boys and girls. Together the two papers argue that average, linear, one-size-fits-all effects are modest, without ruling out that a specific child, a specific app, or a specific pattern of use matters more than the population number suggests. Other researchers, including those behind widely cited cohort studies of depressive symptoms, have used the same kind of survey data to argue for a more worrying pattern; the disagreement is substantially about which analytic choices to trust, not about a different dataset showing a dramatically different world.

A practical next step

Rather than treating a wellbeing statistic as a target to hit, you could use it as a reason to look past total hours toward what is actually happening during screen use in your own home this week.

• Is the number that worries you about total time, or about specific content and company?
• Would a broad wellbeing survey have picked up what changed for your child this month?
• If population averages explain so little, what should that change about applying one household rule to every child?

None of this proves technology use is inconsequential for every teenager. It gives a narrower, more useful fact: at the population level, the measured link between digital technology and adolescent wellbeing is small enough that the method used to look for it matters as much as the number that method returns.