. 2020 Jun 25;15(6):e0233960. doi: 10.1371/journal.pone.0233960

Table 5. Result of scenario 1, a simulation setting that mimics the realistic data, i.e. Subset-BRD dataset.

	P₀	RMSE( ${\hat{P}}_{0}$ )	bias( ${\hat{P}}_{0}$ )	P₁	RMSE( ${\hat{P}}_{1}$ )	bias( ${\hat{P}}_{1}$ )	OR	RMSE( $\hat{O R}$ )	bias( $\hat{O R}$ )
True	0.5931			0.5007			0.6890
UOM	0.6462	0.0557	0.0530	0.4711	0.0308	-0.0296	0.4902	0.2039	-0.1988
MOM	0.6464	0.0559	0.0533	0.4708	0.0311	-0.0298	0.4893	0.2048	-0.1998
IPW	0.5904	0.0284	-0.0027	0.4998	0.0098	-0.0009	0.6980	0.0792	0.0089
ICPW	0.5901	0.0285	-0.0030	0.5030	0.0105	0.0023	0.7081	0.0825	0.0191
DR-IPW	0.5909	0.0228	-0.0022	0.5025	0.0094	0.0018	0.7030	0.0681	0.0140
DR-ICPW	0.5901	0.0230	-0.0030	0.5028	0.0097	0.0021	0.7064	0.0698	0.0174