Skip to main content
. Author manuscript; available in PMC: 2015 Mar 9.
Published in final edited form as: J Biopharm Stat. 2014;24(3):634–648. doi: 10.1080/10543406.2014.888444

Table 3.

Monte Carlo results (based on 500 replicates): Linear regression with missing covariate where N = 200, X1 ~ Poisson(1), X2 ~ N(3 − 0.5 * X1, 1), Y ~ N(b0 + b1 * X1 + b2 * X2, 4), missing rate = 0.37, Spearman correlation coefficient between X1 and X2: −0.42, Spearman correlation coefficient between X1 and Y: 0.40, and Spearman correlation coefficient between X2 and Y: −0.43

b0 = 10.000
b1 = 1.333
b2 = −1.333
Method Esta SDb SEc CRd Est SD SE CR Est SD SE CR
FO 9.921 0.977 0.944 94.6 1.348 0.334 0.320 95.2 −1.307 0.288 0.285 95.6
CC 9.911 1.125 0.995 92.6 1.095 0.348 0.327 87.0 −0.483 0.367 0.339 30.4
IPWDR12 10137 1.367 1.307 93.0 1.239 0.571 0.558 91.0 −1.343 0.377 0.389 95.2
IPWDR21 9.428 1.213 1.174 90.0 1.536 0.454 0.481 90.2 −1.143 0.368 0.360 90.8
IPWDR22 9.543 1.403 1.589 89.6 1.494 0.557 0.758 89.4 −1.188 0.398 0.403 91.4
NNMIB12 M1: misspecified; M2: correctly specified
(3,0.8,0.2)e 10.349 1.152 1.137 92.6 1.155 0.428 0.443 93.4 −1.408 0.334 0.325 92.2
(3,0.5,0.5) 10.264 1.211 1.173 92.6 1.198 0.467 0.464 92.0 −1.387 0.344 0.332 92.2
(3,0.2,0.8) 10.180 1.257 1.209 93.0 1.240 0.501 0.484 93.6 −1.365 0.352 0.341 92.2
(5,0.8,0.2) 10.439 1.113 1.115 93.4 1.112 0.409 0.431 93.8 −1.431 0.325 0.319 92.0
(5,0.5,0.5) 10.324 1.172 1.153 93.6 1.170 0.447 0.456 94.2 −1.402 0.335 0.327 94.2
(5,0.2,0.8) 10.252 1.213 1.193 94.4 1.208 0.480 0.478 93.8 −1.384 0.342 0.335 93.0
NNMIB21 M1: misspecified; M2: misspecified
(3,0.8,0.2) 10.164 1.207 1.167 93.8 1.244 0.459 0.456 93.8 −1.362 0.348 0.355 93.6
(3,0.5,0.5) 10.260 1.168 1.146 92.6 1.197 0.442 0.446 93.2 −1.386 0.338 0.329 93.2
(3,0.2,0.8) 10.373 1.133 1.122 92.6 1.141 0.421 0.432 92.8 −1.414 0.330 0.323 91.0
(5,0.8,0.2) 10.208 1.156 1.144 95.4 1.224 0.437 0.444 93.8 −1.374 0.336 0.329 93.6
(5,0.5,0.5) 10.324 1.134 1.134 94.6 1.165 0.423 0.441 93.8 −1.402 0.330 0.325 93.2
(5,0.2,0.8) 10.474 1.092 1.104 92.8 1.093 0.402 0.424 94.6 −1.439 0.319 0.317 92.2
NNMIB22 M1: misspecified; M2: correctly specified
(3,0.8,0.2) 10.068 1.235 1.211 93.8 1.288 0.474 0.478 92.6 −1.336 0.355 0.345 92.4
(3,0.5,0.5) 10.085 1.243 1.241 94.8 1.283 0.486 0.498 93.4 −1.340 0.354 0.350 92.6
(3,0.2,0.8) 10.084 1.266 1.252 93.6 1.289 0.504 0.506 93.0 −1.341 0.356 0.351 92.4
(5,0.8,0.2) 10.112 1.194 1.176 93.8 1.272 0.455 0.462 93.6 −1.351 0.345 0.336 93.2
(5,0.5,0.5) 10.142 1.212 1.201 94.4 1.261 0.471 0.479 93.0 −1.358 0.346 0.340 93.8
(5,0.2,0.8) 10.157 1.218 1.219 93.8 1.259 0.483 0.491 94.6 −1.362 0.344 0.343 93.0
a

Average of 500 point estimates.

b

Empirical standard deviation of 500 point estimates.

c

Average of 500 estimated standard errors.

d

Coverage rate of 500, 95% confidence intervals.

e

(NN, w1, w2).