When a single correlation of social media use with a positive outcome is found, it immediately becomes headline, while hundreds of other evidence are discarded because they are 'boring'. Well, even this correlation is of little benefit. Anxious people cover their stress while engaging online and postpone their improvement while they are distracted from building real, healthy interpersonal relationships that will help them.

Smoking is correlated with lower risk of Parkinson's disease but it's not suggested in any setting.