Your Kid Is Not a Data Point: Why Differential Susceptibility Changes the Conversation

When a new study comes out connecting social media use to poor teen mental health — and there are many — the coverage tends to follow a predictable pattern. A headline announces the finding in sweeping terms. Parents share it widely. The conclusion that social media is harmful gets a little more entrenched.

What rarely makes it into the coverage is a closer look at what the study actually found, and what it can and can't tell us.

Most of this research is correlational — meaning it identifies relationships between variables, not causes. Maybe you've heard the refrain “correlation is not causation” a million times over. But what gets less attention is that this research is conducted at the population level, which means it describes patterns across large groups of people. The average effect may be small. The variation within the sample — who's most affected, who's least, who might even benefit — often tells a more interesting story than the headline number.

This is where differential susceptibility comes in. It's a concept from developmental science that captures something important: we don't all respond to the same experiences the same way. Biological factors matter. Temperament matters. Context matters enormously. A child who is more emotionally reactive may be more affected — for better and for worse — by a given environment than a child who is less so. The same exposure can mean something very different depending on who's experiencing it and under what circumstances.

Applied to social media, this reframes the question considerably. Rather than asking "is social media bad for teens?" — a question that implies a uniform effect — the more useful questions are: which teens are most vulnerable, and why? What circumstances amplify risk, and what circumstances buffer it? Which platforms and uses carry more risk than others?

The research on these questions is still developing. But what we already know points toward a more nuanced picture than the blanket warnings suggest. Passive use — scrolling, comparing, watching — tends to be associated with worse outcomes than active use, like creating, messaging, connecting. Social media that reinforces existing social relationships looks different from social media that substitutes for them. And the direction of causation is genuinely hard to establish: teens who are already struggling may turn to social media in ways that compound their difficulties, rather than social media causing the difficulties in the first place.

None of this means the concerns are unfounded. There are real risks, and some kids face them acutely. But parenting from the average — or from the headline — means missing what's actually relevant about your specific child. Differential susceptibility is a reminder that knowing your kid, and staying curious about their actual experience, is more protective than any blanket rule.

The question worth sitting with isn't "should my kid be on social media?" It's "how is my particular kid doing with social media, in their particular life, with the particular ways they use it?"

That's harder to answer. But it's the right question.