A famous teen depression and screens study is disputed

The short answer

A well-known 2017 study found that American teenagers who reported more time on new media also reported more depressive symptoms and suicide-related outcomes, especially girls. The association was real in the data the authors used, but the size of that association, and what it can support, has been argued over ever since.

What the evidence says

Twenge, Joiner, Rogers and Martin published their analysis in Clinical Psychological Science, posted online on 14 November 2017 and appearing in the journal's January 2018 issue. The study combined two nationally representative surveys of US adolescents in grades 8 to 12, covering 506,820 respondents, with separate national suicide statistics for ages 13 to 18. It reported that depressive symptoms and suicide-related outcomes rose between 2010 and 2015, alongside a rise in time spent on smartphones and other new media, while time spent on non-screen activities moved the other way. The authors noted that economic indicators did not track the same trend over that period, which they treated as evidence against a purely economic explanation.

For context

The design is correlational and cross-sectional in its screen-time measure: it cannot show that media use caused the rise in symptoms rather than tracking alongside it. The dispute that followed centres on effect size. Orben and Przybylski's 2019 specification-curve analysis, applied to the same style of large adolescent survey data, found that technology-use measures explained a very small share of wellbeing variation once many reasonable analytic choices were tested rather than one preferred model. Their critique is not that the correlation Twenge and colleagues reported is fabricated, but that a single analytic path can make a small, uncertain association look like a clear trend line, and that specification choices deserve as much scrutiny as the headline number.

A practical next step

Instead of treating this study as proof that a rising graph line explains an individual teenager's mood, you could use it as a prompt to ask what is changing in your teenager's specific pattern of use and sleep around the same period the data describes.

• Is the concern about total new-media time, or about a specific change in your teenager's routine?
• What non-screen activities dropped out of the week when screen use went up?
• Would a mood change this year show up in a national survey average at all?

Both the original finding and its critics are working from real, published data. The honest summary is that a national rise in depressive symptoms coincided with a national rise in new-media use among US teenagers between 2010 and 2015, that the association is disputed on grounds of size and method rather than existence, and that neither side has shown the coincidence is causal.