A pairwise pseudo-likelihood approach for regression analysis of left-truncated failure time data with various types of censoring

Table 3 Simulation results for the comparison of the proposed method with Wu et al. (2018)’s method under right censored data, including the estimated bias (Bias), the sample standard error (SSE) of the estimates, the average of the standard error estimates (SEE), and the 95% empirical coverage probability (CP)

		Proposed method					Wu et al. (2018)’s method
n	Par	True	Bias	SSE	SEE	CP	Bias	SSE	SEE	CP
\(A^{*}\) follows the uniform distribution
100	\(\beta _1\)	1	0.025	0.244	0.247	95.0	0.026	0.244	0.227	92.8
	\(\beta _2\)	1	0.027	0.391	0.398	94.9	0.027	0.391	0.368	93.6
300	\(\beta _1\)	1	0.011	0.129	0.133	96.0	0.012	0.129	0.130	95.2
	\(\beta _2\)	1	0.005	0.246	0.216	94.8	0.005	0.216	0.211	95.1
500	\(\beta _1\)	1	0.005	0.100	0.102	95.2	0.005	0.100	0.100	95.1
	\(\beta _2\)	1	0.002	0.166	0.165	95.1	0.003	0.166	0.162	94.9
\(A^{*}\) follows the exponential distribution
100	\(\beta _1\)	1	0.024	0.248	0.257	95.8	0.024	0.248	0.237	94.3
	\(\beta _2\)	1	0.015	0.398	0.416	95.7	0.015	0.398	0.383	93.6
300	\(\beta _1\)	1	0.003	0.134	0.138	95.6	0.003	0.134	0.135	95.3
	\(\beta _2\)	1	0.008	0.218	0.223	95.4	0.008	0.218	0.219	95.2
500	\(\beta _1\)	1	0.010	0.107	0.106	94.7	0.010	0.107	0.105	95.2
	\(\beta _2\)	1	0.011	0.172	0.171	94.8	0.011	0.172	0.169	94.6

ISSN: 1471-2288