Research Article

Robust Wild Bootstrap for Stabilizing the Variance of Parameter Estimates in Heteroscedastic Regression Models in the Presence of Outliers

Table 1

Wild bootstrap standard errors of the parameters for the Concrete Compressive Strength data.

Standard error (Se)Classical wild bootstrapRobust wild bootstrap

BootolsBootwuBootliuRBootolsRBootwuRBootliu

Data with outliers
 Intercept29.003727.067019.455927.195519.234316.5458
 Cement0.008990.008740.006120.008570.006410.00503
 Blast Furnace Slag0.010900.010220.007400.010470.007570.00605
 Fly Ash0.012830.012960.008640.011800.009710.00694
 Water0.045840.041290.029140.043010.029860.02566
 Superplasticizer0.098240.095640.066400.102530.071040.05706
 Coarse Aggregate0.010190.009560.006870.009280.006660.00559
 Fine Aggregate0.011130.010560.007720.010620.007720.00653
 Age0.007960.005500.005560.003010.004190.00175
 Average Se3.245533.029052.177083.043872.153051.85116

Data without outliers
 Intercept26.394420.648317.558225.583820.736716.9330
 Cement0.008560.006870.005540.008100.006660.00531
 Blast Furnace Slag0.010000.008050.006790.009840.007850.00648
 Fly Ash0.011880.010140.007570.011430.010220.00736
 Water0.040680.030510.026460.038370.030800.02580
 Superplasticizer0.093840.071510.059400.094980.071460.05974
 Coarse Aggregate0.009090.007520.006020.009030.007390.00575
 Fine Aggregate0.010470.008230.006990.010250.008260.00679
 Age0.009920.008420.006900.010160.008370.00619
Average Se 2.954312.311061.964872.864002.320861.89516