Table 2 Means and standard errors (in parentheses) of performance measures in experiment I.

From: Estimating covariate-balanced survival curve in distributed data environment using data collaboration quasi-experiment

Method

Sample size after matching

MASMD

Inconsistency

Gap(T)

Gap(C)

\(\:{D}^{\text{C}\text{A}}\)

\(\:{D}^{\text{t}\text{r}\text{u}\text{e}}\)

LA

370.18 (20.71)

0.6820 (0.1059)

0.1705 (0.014)

0.0430 (0.0159)

0.0547 (0.0196)

0.1096 (0.0526)

0.1656 (0.0686)

LMCA (L-clb)

739.84 (28.66)

0.6720 (0.0791)

0.1792 (0.0149)

0.0322 (0.0133)

0.0477 (0.0149)

0.1036 (0.0416)

0.1632 (0.047)

DC-QE (T-clb)

310.16 (21.88)

0.1847 (0.0508)

0.0480 (0.0124)

0.0382 (0.0156)

0.0396 (0.0145)

0.0530 (0.0409)

0.0886 (0.0609)

DC-QE (L-clb)

690.20 (74.26)

0.3404 (0.1538)

0.1033 (0.0518)

0.0262 (0.0137)

0.0331 (0.0152)

0.0526 (0.0408)

0.0998 (0.0576)

DC-QE (W-clb)

668.18 (51.51)

0.2772 (0.1210)

0.0857 (0.0404)

0.0259 (0.0111)

0.0275 (0.0121)

0.0425 (0.0315)

0.0854 (0.0502)

CA (reference)

624.58 (29.81)

0.1211 (0.0229)

0.0000 (0.0000)

0.0000 (0.0000)

0.0000 (0.0000)

0.0000 (0.0000)

0.0690 (0.0449)

  1. (Gap(C): Gap for control group, Gap(T): Gap for treatment group, \(\:{D}^{\text{C}\text{A}}\): Difference of RMST difference relative to CA, \(\:{D}^{\text{t}\text{r}\text{u}\text{e}}\): Difference of RMST difference relative to the true survival function, LA: local analysis, LMCA: local matching and central analysis, CA: central analysis. The best result is bold, and the second-best result is underlined).