. 2024 Jun 3;24:125. doi: 10.1186/s12874-024-02254-x

Table 1.

TPR and FDP for the four true mediators (M1–M4)

Settings	MT methods^a	CONF methods^b*	TPR				Overall TPR	FDP
Settings	MT methods^a	CONF methods^b*	M1 $α β$ = 0.16	M2 $α β$ = 0.20	M3 $α β$ = 0.25	M4 $α β$ = 0.30	Overall TPR	FDP
n = 300, p = 1000	HIMA	RA	0.2340	0.2740	0.4820	0.4880	0.3695	0.0027
		PSR	0.1980	0.2260	0.4380	0.4340	0.3240	0.0031
		IPW	0.2000	0.2060	0.3760	0.4120	0.2985	0.0050
		OW	0.1860	0.2000	0.4140	0.4080	0.3020	0.0033
	mHIMA2	RA	0.3298	0.3510	0.6004	0.5962	0.4693	0.0144
		PSR	0.4754	0.5246	0.7295	0.7520	0.6204	0.0670
		IPW	0.5071	0.5253	0.7333	0.7394	0.6263	0.0909
		OW	0.5396	0.5375	0.7546	0.7789	0.6526	0.0905
n = 300, p = 10,000	HIMA	RA	0.2660	0.2680	0.4540	0.4940	0.3705	0.0185
		PSR	0.2160	0.2280	0.4040	0.4460	0.3235	0.0046
		IPW	0.2020	0.1820	0.3640	0.4080	0.2890	0.0120
		OW	0.1960	0.2100	0.3840	0.4180	0.3020	0.0082
	mHIMA2	RA	0.3550	0.3203	0.5649	0.5758	0.4540	0.0141
		PSR	0.4714	0.4857	0.6776	0.6918	0.5816	0.0539
		IPW	0.5051	0.5030	0.6707	0.7131	0.5980	0.0800
		OW	0.5466	0.5344	0.7328	0.7672	0.6452	0.0834
n = 500, p = 1000	HIMA	RA	0.5440	0.5860	0.8380	0.8740	0.7105	0.0007
		PSR	0.5140	0.5420	0.8160	0.8480	0.6800	0.0022
		IPW	0.4540	0.4480	0.7280	0.7420	0.5930	0.0025
		OW	0.4800	0.4880	0.7680	0.7880	0.6310	0.0024
	mHIMA2	RA	0.6579	0.6984	0.8927	0.9170	0.7915	0.0076
		PSR	0.8060	0.8180	0.9400	0.9540	0.8795	0.0487
		IPW	0.8377	0.8297	0.9198	0.9599	0.8868	0.0679
		OW	0.8620	0.8520	0.9480	0.9760	0.9095	0.0629
n = 500, p = 10,000	HIMA	RA	0.5440	0.5640	0.8460	0.8360	0.6975	0.0050
		PSR	0.5160	0.5360	0.8220	0.8020	0.6690	0.0000
		IPW	0.4140	0.4480	0.7280	0.7220	0.5780	0.0026
		OW	0.4520	0.4900	0.7920	0.7720	0.6265	0.0016
	mHIMA2	RA	0.6451	0.6326	0.8852	0.8664	0.7573	0.0109
		PSR	0.8116	0.8196	0.9519	0.9479	0.8828	0.0657
		IPW	0.8096	0.8236	0.9419	0.9339	0.8773	0.0774
		OW	0.8417	0.8717	0.9760	0.9559	0.9113	0.0667

^a MT methods denote two different mediation test approaches, including HIMA Zhang et al. and modified HIMA2 Perera et al. (termed mHIMA2)

^b CONF methods denote different confounding adjustment methods

^* RA denotes regression adjustment. PSR denotes propensity score regression adjustment. IPW denotes inverse probability weighting. OW denotes overlapping weighting