Skip to main content
. 2017 Oct 26;104(4):881–899. doi: 10.1093/biomet/asx056

Table 2.

Simulation results for Case 2A: the two-stage instrumental variable estimator and the naïve rank estimator; all values have been multiplied by Inline graphic

Inline graphic Method Parameter Bias ESE ASE ECR
100 Two-stage Inline graphic –2.5 24.5 25.4 94.6
    Inline graphic 0.2 30.4 29.8 95.4
  Naïve Inline graphic 28.9 10.1 10.1 18.6
    Inline graphic –28.8 18.1 17.6 62.2
200 Two-stage Inline graphic –2.5 17.7 17.4 91.7
    Inline graphic 0.6 20.8 20.3 93.3
  Naïve Inline graphic 28.5 6.8 7.1 1.2
    Inline graphic –28.1 12.3 12.3 37.8
400 Two-stage Inline graphic –0.3 11.7 12.3 95.8
    Inline graphic –0.5 13.7 14.2 97.0
  Naïve Inline graphic 27.9 4.7 5.0 0.0
    Inline graphic –27.7 8.8 8.6 13.0

Bias, average bias; ESE, empirical standard error; ASE, average of the estimated standard errors; ECR, empirical coverage rate of the 95% Wald confidence intervals.