Skip to main content
. Author manuscript; available in PMC: 2013 May 29.
Published in final edited form as: Biometrika. 2006 Dec;93(4):763–775. doi: 10.1093/biomet/93.4.763

Table 1.

Simulation results for nonparametric estimators under the assumptions of independent and informative observation times

(a) Independent observation times
n t Λ0(t) Proposed NPMLE NPMPLE
Bias SE Bias SE Bias SE
100 1·0 2 −0·031 0·397 −0·030 0·392 −0·060 0·418
3·0 6 −0·006 0·488 −0·019 0·476 −0·058 0·633
5·0 10 −0·024 0·542 −0·047 0·505 −0·063 0·723
7·0 14 −0·015 0·654 −0·055 0·559 −0·065 0·844
9·0 18 0·004 0·804 −0·051 0·643 −0·045 0·904
1000 1·0 2 0·011 0·156 0·010 0·154 0·007 0·200
3·0 6 0·005 0·202 −0·002 0·189 −0·017 0·276
5·0 10 0·014 0·210 −0·002 0·193 −0·030 0·336
7·0 14 0·011 0·241 −0·009 0·211 −0·019 0·374
9·0 18 0·025 0·277 −0·003 0·218 −0·023 0·408
(b) Informative observation times
n t Λ0(t) Proposed NPMLE NPMPLE
Bias SE Bias SE Bias SE
100 1·0 2 −0·009 0·271 0·015 0·271 0·447 0·454
3·0 6 −0·007 0·553 −0·040 0·512 0·361 0·795
5·0 10 0·049 0·874 −0·543 0·722 −1·183 0·949
7·0 14 0·008 1·198 −1·699 0·859 −3·395 1·067
9·0 18 0·127 1·532 −2·987 1·021 −5·324 1·349
1000 1·0 2 0·000 0·104 0·021 0·101 0·484 0·194
3·0 6 0·015 0·214 −0·021 0·182 0·431 0·345
5·0 10 0·031 0·343 −0·509 0·258 −1·157 0·407
7·0 14 0·035 0·459 −1·580 0·309 −3·321 0·459
9·0 18 0·038 0·564 −2·901 0·340 −5·448 0·508

NPMLE, nonparametric maximum likelihood estimator; NPMPLE, nonparametric maximum pseudolikelihood estimator; Bias and SE are the Monte Carlo sample mean and standard deviation of the 1000 estimates of Λ0(t).