A study finds a strong positive association between screen time and anxiety score. The participants chose whether to take part and both variables were self-reported.
Give three reasons why a causal conclusion is not justified.
Describe an improved observational study.