Fig. 2.

Comparison of adjusted 2SLS IV model estimates from 200 simulated data sets. Different strengths of IV (α₁ = 0.1, 0.3, 0.5) are given across the x-axis (a-c). Different strength of treatment effect (β₁ = 1.0, 3.0) and sample sizes (N = 20,000, 200,000) are given across the y-axis (1–4); dashed blue horizontal line is the true causal treatment effect β₁ =3.0, error bars are 95% confidence intervals for the treatment effect estimates for each simulated data set. Black error bars indicate a significant effect estimate, red error bars indicate a non-significant effect estimate