We cannot simply ban things if they are bad, however weak or strong the evidence is. We have to examine access on the basis of rights alone. If the right exists then it could kill all kids and it would still be the correct decision to allow it.
With subjects like this, the outcome variables are things like “suicide” and “eating disorders”. So, what, we have to wait until R² > .95 to be convinced of the utility regulation? Seems the least-harm approach is actually to index on the side of regulation first instead of pretending the null is true in this case.
> The Science is Not Settled: How Weak Evidence is Fueling a National Push to Ban Social Media for Youth