Observational studies. Problem with self selection (identify kids who decided to watch porn in the past, and compare with those that didn't) probably will not

Observational studies. Problem with self selection (identify kids who decided to watch porn in the past, and compare with those that didn't) probably will not be that huge, when the people are only outraged about "kids" self selecting to watch porn.