Again, its hard to separate that phenomenon from potentially confounding effects. Is the fact that children are typically hurt by people they know the result of parents being reluctant to let their kids out on the street by themselves, thereby limiting the opportunities for potential strangers to harm them?

Again, the effect precedes the behavior change, so is unlikely to be a result of it.