Table 2.

Simulation results when Gaussian random effects were assumed for the bimodal random effects but t_i in (2.2) was varied. The simulation included 500 data sets with 500 subjects each

	Time points	(b0i)	(b1i)	β0	(b0i)	(b0i, b1i)	(b1i)
Truth		5.750	– 0.600	0.500	0.360	– 0.228	0.903
MC Avg	tA	5.718	– 0.522	0.494	0.341	– 0.180	0.732
	tB	5.715	– 0.520	0.496	0.337	– 0.175	0.722
	tC	5.757	– 0.590	0.500	0.360	– 0.234	0.886
MC SD	tA	0.045	0.048	0.058	0.035	0.036	0.067
	tB	0.046	0.048	0.059	0.033	0.034	0.066
	tC	0.040	0.044	0.054	0.027	0.029	0.049
Avg SE	tA	0.044	0.042	0.059	0.034	0.033	0.062
	tB	0.043	0.041	0.058	0.032	0.032	0.061
	tC	0.039	0.043	0.053	0.026	0.030	0.058
CP	tA	0.878	0.532	0.956	0.896	0.662	0.268
	tB	0.864	0.474	0.950	0.864	0.586	0.212
	tC	0.944	0.930	0.948	0.938	0.962	0.960

t_A = (0, 1, 2, 3, 4)^T; t_B = (0, 1, 2, 3, 3.5, 4, 4.5)^T; t_C = (– 2, – 1, 0, 1, 2, 3, 4)^T; MC Avg, Monte Carlo average of the parameter estimates; MC SD, Monte Carlo standard deviation of the parameter estimates; Avg SE, average of the standard error estimates; CP, Monte Carlo coverage probability of the 95% Wald-type confidence intervals.