This series has just published five articles that contradict each other, and that was the plan.
Jonathan Haidt argued that a childhood was rewired and it cost a generation its mental health. Jean Twenge showed the curves that bend in 2012. Sarah Coyne followed five hundred young people for eight years and found that when a given teenager's own social media use went up, her own depression did not. Eva Telzer found that habitual checking at twelve changed how a brain develops, and then declined to call that harm.
If you are looking for which one of them is right, you are asking the question this field has spent a decade proving is badly formed. Jacqueline Nesi is the researcher who most clearly reformulated it.
Nesi is a clinical psychologist and researcher at Brown University's Warren Alpert Medical School, in the Department of Psychiatry and Human Behavior, and she also practices at Rhode Island Hospital, which matters for what follows. Her research, funded through an NIMH career development award, does not ask whether social media harms adolescents. It asks which adolescents, and through what route.
That sounds like a small adjustment. It is the whole argument.
Consider why the big studies keep disagreeing. If you take ten thousand teenagers, measure their social media use and their depression, and compute an average association, you will get a number near zero. Both camps then use it. The skeptics say: look, the effect is tiny, smaller than things nobody worries about. The alarm-raisers say: your average is hiding the fire.
They are both describing the same arithmetic correctly. An average near zero is exactly what you get when a thing helps some people, harms others, and does nothing to most. The near-zero average is not evidence of no effect. It is evidence of heterogeneity, and it is the single most robust finding in this literature.
Which means the useful questions were never the ones being argued about on television. Not whether it is bad. Which teenager, which platform behavior, which existing vulnerability, which outcome. Nesi's work is on that: identifying the adolescents for whom social media use genuinely tracks depression and suicidality risk, rather than issuing a verdict that is wrong for most of the population in one direction or the other.
Her second contribution is the one I would hold up to anyone building anything in this space, and it is why she anchors the sixth book in The Unmistakably You Series, Mean It.
Since 2022 she has written Techno Sapiens, a newsletter that takes the week's new research on adolescents and technology and explains it honestly to parents. Honestly includes the weeks when the new study undercuts what she said in a previous one. It includes the weeks when the answer is that nobody knows. There is no academic credit for this. It is not a publication, it does not count for tenure, and nobody is grading her on it.
That is what conviction actually looks like in practice, and it is why Mean It is about examined positions rather than loud ones. It is easy to hold a strong line in a contested field: pick a side, repeat it, and let the disconfirming studies pass unmentioned. It is much harder to tell the same audience, week after week, what the evidence currently supports, including when that has moved.
So here is the standard this sequence was built to meet, and the reason it contains five pieces that cannot all be right.
If you only publish the researchers who agree with your thesis, you do not have a thesis. You have a marketing position with citations attached. Teens UnEdited claims that adolescents edit themselves down and that belonging protects them. Coyne's eight-year null is the most inconvenient result in this series for a simple screens story, and it is in the set, stated at full strength, because the claim this series actually makes survives it. Haidt's position is in the set too, stated at full strength, with its critics named.
There is a reason she can hold that line, and it is not temperament. She sees patients. Alongside the research she practices clinically at Rhode Island Hospital, which means the adolescents in her data are not an abstraction to her and neither are the ones whose parents read her newsletter. A researcher who only reads datasets can afford a tidy position. A clinician cannot, because the next young person through the door will be one of the ones the average got wrong. She has also co-developed a parent-facing course with Emily Weinstein, whose work we featured in The Problem Was Never the Phone, which tells you something about how small and how serious the group of people doing this translation properly actually is.
Nesi's question is the one that makes a parent's job tractable. Not whether the internet is bad for teenagers. Whether it is bad for this teenager, in this way, given what she already carries.
That question has an answer you can actually find. The other one never did.
Want to explore these ideas further? Learn more about my work on leadership, identity and conviction at www.theglennllopis.com.
This essay was first published on Conviction Chronicles

