**Causation vs. correlation.** Most studies are correlational. It is possible that the causal arrow runs in the opposite direction: adolescents who are already depressed or anxious may use social media more as a coping mechanism. Longitudinal studies offer some evidence of bidirectional effects, but