It's pretty hard to study beyond doubt though. You're not going to get much parental consent (or probably even institutional approval) to run a study where you

It's pretty hard to study beyond doubt though. You're not going to get much parental consent (or probably even institutional approval) to run a study where you require the experimental group of teenagers to spend 3+ hours a day on social media. And if the experimental group limits screentime, well I think parents of teens know how hard that can be to actually enforce across devices if the kid really wants to get on social media. You can't control what their peer groups are doing, and there are privacy limitations on what info you can even collect on specific phone activity from consenting adults, let alone teens. In general, for studies of difficult long term behavior adjustments you basically need cooperative participants that will be honest when they slipped up. Teens are a tough population for this.