TABLE 1.

Randomized study results under Model 1 scheme 1, with $\pi =0.5,\alpha =0.05,n=1000$

c	Method	${\mu }_{\text{IPW}}$	${\mu }_{\text{AIPW}}$	${\sigma }_{\text{IPW}}$	${\sigma }_{\text{AIPW}}$	$1-{\beta }_{\text{IPW}}/{\alpha }_{\text{IPW}}$	$1-{\beta }_{\text{AIPW}}/{\alpha }_{\text{AIPW}}$
0.30	SAP-match	1.61	2.50	1.01	1.01	0.51	0.80
0.30	SBT	0.62	2.55	1.00	0.99	0.15	0.81
0.40	SAP-match	2.29	3.39	1.01	1.00	0.73	0.96
0.40	SBT	0.92	3.44	1.00	1.01	0.21	0.97
0.50	SAP-match	2.92	4.40	0.94	1.02	0.92	1.00
0.50	SBT	1.17	4.43	1.00	1.01	0.30	1.00
0.00	SAP-match	0.03	−0.03	1.00	0.99	0.04	0.05
0.00	SBT	−0.03	0.02	1.01	1.00	0.06	0.06
−1.00	SAP-match	0.02	−0.06	0.99	1.00	0.04	0.05
−1.00	SBT	−0.05	−0.07	1.02	1.04	0.06	0.05
−2.00	SAP-match	0.03	−0.03	0.98	0.99	0.05	0.05
−2.00	SBT	−0.04	−0.00	1.02	1.03	0.06	0.06

Note:Mean ($\mu$), standard deviation ($\sigma$), and power or type I error ($1-\beta /\alpha$) of the one-step value difference test statistic based on 500 simulated data sets. Subscripts “IPW” and “AIPW” correspond to $T_{\text{IPW},•}$ and $T_{\text{AIPW},•}$, respectively. “Method” refers to chunking method. Largest standard error for $\mu$, $\sigma$, and $1-\beta /\alpha$ is 0.05, 0.07, and 0.02, respectively.