Research Article

Bayesian Adaptive Lasso for Regression Models with Nonignorable Missing Responses

Table 6

Bayesian estimates for type i in the example 3.

ParameterTrueMeanMedian5%95%SDRMSE

(i) Empirical Bayesian scenario
3.03.0052.9842.8423.1920.1110.111
1.51.5021.5091.3051.6390.1110.111
0.00.0040.000−0.1000.1200.0770.077
0.0−0.004−0.001−0.1070.0720.0610.061
2.01.9781.9881.8152.1290.1010.103
0.00.0140.001−0.0390.0910.0540.056
0.00.0020.000−0.1000.1120.0570.057
0.00.0050.001−0.0620.0970.0580.058
1.00.9940.9890.8871.0810.0590.059
−1.0−1.087−1.089−1.383−0.8190.1740.193
0.30.3660.3840.0870.5930.1590.171
0.30.3360.324−0.4100.9550.4410.440
0.30.2580.292−0.4480.7590.3720.371
0.30.3570.3290.0380.6670.2110.217
0.00.027−0.007−0.3800.4370.2600.260
0.0−0.066−0.098−0.6770.4780.3370.342
0.00.013−0.015−0.4560.3110.2490.248
0.0−0.018−0.056−0.4870.4100.2820.281
0.0−0.015−0.004−0.3930.3480.2470.246

(ii) Hierarchical model scenario
3.03.0103.0012.8133.1850.1200.120
1.51.4681.4571.2821.6330.1100.114
0.00.0280.018−0.1480.1650.1030.106
0.0−0.0040.002−0.1910.1500.1090.108
2.01.9911.9691.8122.2180.1260.126
0.0−0.003−0.001−0.1780.1420.1020.102
0.0−0.007−0.008−0.1610.1400.1040.103
0.00.002−0.007−0.1570.1650.0940.093
1.01.0191.0200.9081.1080.0630.066
−1.0−1.052−1.055−1.290−0.8490.1430.151
0.30.3190.3000.0310.5810.1750.175
0.30.3790.351−0.4941.0810.5030.507
0.30.3710.364−0.2290.9070.3380.344
0.30.3430.314−0.0820.7090.2370.240
0.00.0210.012−0.3810.4260.2550.255
0.00.0200.006−0.6010.5660.3720.371
0.00.0320.038−0.4970.3990.2720.271
0.0−0.054−0.071−0.5120.3510.2700.275
0.00.0740.065−0.2850.4130.2170.229