Skip to main content
. Author manuscript; available in PMC: 2013 Jan 1.
Published in final edited form as: Stat Sin. 2012;22:149–172.

Table 2.

Simulation results when true models are Model (O1) for Y and Model (M2) for δ, with μ = 10, and n = 400.

Method RB(%) SD SE MSE CR(%)

CC 17.47 0.256 0.259 3.119 0.0
Correct Working Models for Both
CE -0.6 1.640 0.515 0.269 91.7
PMI 0.04 0.323 0.241 0.058 87.0
MIB(3,1.0,0.0) 1.80 0.365 0.375 0.173 90.0
MIB(3,0.8,0.2) 1.83 0.396 0.375 0.174 88.6
MIB(3,0.5,0.5) 2.04 0.420 0.396 0.198 89.3
MIB(3,0.2,0.8) 2.40 0.439 0.420 0.234 88.8
MIB(3,0.0,1.0) 2.99 0.462 0.487 0.327 90.3

Wrong Working Model for Y only (O1W)
CE -0.57 3.659 0.759 0.579 79.9
PMI 9.15 0.302 0.234 0.892 7.1
MIB(3,1.0,0.0) 9.92 0.314 0.327 1.091 24.5
MIB(3,0.8,0.2) 4.30 0.366 0.354 0.310 77.8
MIB(3,0.5,0.5) 3.61 0.404 0.382 0.276 82.6
MIB(3,0.2,0.8) 3.13 0.430 0.408 0.264 86.3
MIB(3,0.0,1.0) 2.97 0.463 0.488 0.328 90.6

Wrong Working Model for δ only (M2W)
CE 0.02 0.377 0.334 0.112 92.6
MIB(3,1.0,0.0) 1.78 0.366 0.374 0.172 91.9
MIB(3,0.8,0.2) 2.30 0.381 0.355 0.179 87.8
MIB(3,0.5,0.5) 3.09 0.392 0.359 0.224 84.7
MIB(3,0.2,0.8) 4.23 0.387 0.362 0.310 78.6
MIB(3,0.0,1.0) 10.72 0.373 0.373 1.288 31.5