Data Collectionhard
0:00.0

A researcher studies the causal effect of worker education (E) on firm productivity (Y). The causal diagram includes:

  • Worker ability (A) → E and A → Y (unmeasured confounder)
  • Prior work experience (W) → E and W → Y (measured)
  • Industry sector (I) → E and I → Y (measured)
  • Education → Productivity (direct causal effect of interest)

To satisfy the backdoor criterion and estimate the causal effect of E on Y, which variable sets are sufficient for adjustment?