Table 4.

Properties and performance of the linear quadrature, optimal in the minimax sense, compared to the optimal trapezoid and GL or CC approaches for Model 3

$n$	$c_v[\%]$	Method	$\sqrt{\mathrm{Objective}}$	Bias	RMSRE	Maximum deviance	Maximum relative deviance
6	10	optimal	4.53E-2	0.787	4.30E−2	32.7	.1712
		opt. trap.	6.01E−2	−0.602	4.74E−2	−31.2	.2071
		GL	5.74E−2	−0.329	4.89E−2	−36.9	−.1995
	5	optimal	2.28E−2	0.312	2.22E−2	15.1	.0960
		opt. trap.	3.81E−2	−0.197	2.52E−2	13.8	.1122
		GL	3.32E−2	−0.334	2.48E−2	−15.3	.1114
	0	optimal	1.35E−3	0.027	6.24E−4	0.29	−.0013
		opt. trap.	3.33E−2	−0.439	1.34E−2	3.16	.0328
		GL	1.96E−2	−0.333	5.06E−3	−0.99	−.0190
4	10	optimal	5.47E−2	0.810	5.30E−2	−44.3	−.2187
		opt. trap.	7.80E−2	−1.105	5.93E−2	−40.4	.2808
		CC	9.17E−2	1.937	6.49E−2	40.2	.2897
	5	optimal	2.84E−2	0.084	2.79E−2	20.1	.1083
		opt. trap.	5.32E−2	−0.986	3.29E−2	−21.0	.1346
		CC	6.69E−2	1.849	3.27E−2	20.7	.1616
	0	optimal	5.72E−3	0.024	2.47E−3	1.22	.0057
		opt. trap.	4.18E−2	−0.833	1.59E−2	4.44	.0417
		CC	5.62E−2	1.890	2.21E−2	3.34	.0551
2	10	optimal	0.104	2.573	0.083	57.5	.3339
		opt. trap.	0.120	−0.503	0.095	−61.1	.4148
		GL	0.227	−8.866	0.122	−55.2	−.4542
	5	optimal	0.075	1.948	0.046	33.6	.1917
		opt. trap.	0.085	−0.130	0.052	−28.6	.2180
		GL	0.218	3.908	0.104	−29.3	.3566
	0	optimal	0.062	−1.174	0.021	6.76	.0603
		opt. trap.	0.070	−0.042	0.026	−8.22	.0663
		GL	0.215	−8.841	0.097	−13.97	−.2121