Observational study

RCT를 할 수 없는 이유들¹

• Ethical reasons (e.g., unethical to randomize people to smoke for measuring effect on lung cancer)
• Infeasibility (e.g., can’t randomize countries into communist/capitalist systems to measure effect on GDP)
• Impossibility (e.g., can’t change a living person’s DNA at birth for measuring effect on breast cancer)

Footnotes

1. A brief introduction to casual inference ↩