Research Article

Bayesian Adaptive Lasso for Regression Models with Nonignorable Missing Responses

Table 3

Bayesian estimates for case iii in the first simulation study.

ParameterTrueMeanMedian5%95%SDRMSE

(i) Empirical Bayesian scenario
5.05.0034.9924.8265.1850.1110.111
0.0−0.011−0.000−0.1530.0460.0680.068
0.00.003−0.001−0.0600.0530.0540.054
0.00.012−0.001−0.1930.0850.0760.076
0.00.0080.001−0.0780.0910.0650.066
0.00.0070.001−0.0830.0820.1110.111
0.00.0040.001−0.1070.0800.0530.052
0.00.0000.000−0.1190.0780.0640.063
1.00.9910.9840.8851.0800.0580.058
−1.0−1.065−1.071−1.310−0.7980.1630.175
0.30.3590.3610.0980.5290.1230.136
0.30.2730.243−0.4420.9580.4430.442
0.30.3230.288−0.0730.7460.2850.285
0.30.3670.3240.0060.7450.2440.252
0.30.3180.307−0.1710.7100.2620.262
0.30.3460.296−0.1090.8220.2960.298
0.30.3220.339−0.1870.7830.2970.296
0.30.3260.300−0.0230.6680.2280.229
0.30.3420.341−0.0380.7220.2340.237

(ii) Hierarchical model scenario
5.05.0095.0064.8755.1630.0910.091
0.0−0.0000.001−0.0580.0370.0380.038
0.0−0.001−0.002−0.0460.0350.0310.031
0.00.004−0.000−0.0420.0480.0290.029
0.00.0020.002−0.0500.0630.0360.036
0.0−0.0010.000−0.0580.0370.0380.038
0.00.0060.005−0.0430.0550.0300.030
0.00.0010.003−0.0440.0430.0290.029
1.01.0131.0090.9271.1060.0600.061
−1.0−1.089−1.085−1.348−0.8610.1500.174
0.30.3520.3480.1490.5150.1180.129
0.30.2890.323−0.4831.0310.4690.467
0.30.3480.338−0.2120.7890.2810.283
0.30.3690.362−0.0620.7350.2460.254
0.30.2830.256−0.1760.7310.2840.283
0.30.3530.343−0.0850.7690.2600.264
0.30.3580.326−0.1290.7700.2620.267
0.30.3640.352−0.1470.7590.2800.286
0.30.2920.270−0.1170.7390.2610.260