Table 2. Root mean-squared error (RMSE) and bias of estimators of the sizes N of simulated populations.

N	Performance measure	Expected Pr(encounter)1	Number of encounters or lists
			2		3			4			5
			LLM-AIC	BLCM	LLM-AIC	LLM-BMA	BLCM	LLM-AIC	LLM-BMA	BLCM	LLM-AIC	LLM-BMA	BLCM
1,000	RMSE	0.025	> 109	692	> 109	543	506	> 109	557	423	> 109	618	376
		0.050	> 109	404	> 109	716	390	> 109	633	424	> 109	572	417
		0.100	> 109	491	> 109	630	541	> 109	531	478	2,005	466	432
		0.150	> 109	566	> 109	537	502	> 109	452	433	276	396	381
		0.200	> 109	508	> 109	464	447	> 109	389	373	190	335	322
	Bias	0.025	> 109	-680	> 109	-31	-468	< −109	44	-356	> 109	109	-279
		0.050	> 109	-307	> 109	220	-42	> 109	194	26	> 109	175	56
		0.100	> 109	55	> 109	251	179	> 109	203	163	-22	173	148
		0.150	> 109	135	> 109	203	182	> 109	162	158	-38	136	138
		0.200	> 109	120	< −109	165	164	> 109	128	139	-49	102	115
10,000	RMSE	0.025	> 109	4,215	> 109	7,386	2,874	> 109	6,303	4,316	> 109	5,794	5,334
		0.050	> 109	3,533	> 109	6,173	5,626	43,319	5,551	5,355	> 109	5,294	5,155
		0.100	> 109	5,336	11,655	5,318	5,246	2,872	4,860	4,777	2,092	4,520	4,423
		0.150	> 109	4,953	3,603	4,766	4,701	2,002	4,235	4,179	1,544	3,798	3,751
		0.200	> 109	3,929	2,511	4,290	3,432	1,479	3,662	2,948	1,178	3,141	2,583
	Bias	0.025	> 109	-3,932	> 109	2,888	28	> 109	2,496	1,644	> 109	2,333	2,457
		0.050	> 109	117	> 109	2,663	2,275	-391	2,386	2,229	> 109	2,264	2,171
		0.100	> 109	1,733	71	2,265	2,290	-269	2,001	2,091	-321	1,820	1,945
		0.150	> 109	1,557	-19	1,901	2,084	-173	1,685	1,880	-326	1,475	1,691
		0.200	> 109	182	-6	1,639	829	-294	1,407	763	-475	1,162	643
20,000	RMSE	0.025	> 109	8,679	> 109	13,859	6,127	57,543	12,382	9,052	31,361	11,618	10,649
		0.050	> 109	7,622	> 109	11,758	11,443	52,505	11,007	10,828	8,080	10,548	10,418
		0.100	> 109	10,366	19,729	10,395	10,300	4,825	9,640	9,480	3,736	8,971	8,791
		0.150	> 109	9,731	5,732	9,413	9,326	3,285	8,345	8,294	2,650	7,490	7,489
		0.200	> 109	9,188	4,558	8,473	8,479	2,646	7,191	7,274	2,270	6,112	6,254
	Bias	0.025	> 109	-7,897	> 109	5,791	524	-613	5,277	3,121	1,275	4,964	4,267
		0.050	> 109	1,214	> 109	5,284	5,035	-836	4,883	4,781	-194	4,621	4,598
		0.100	> 109	3,447	574	4,434	4,643	-188	4,012	4,308	-388	3,670	4,065
		0.150	> 109	3,117	50	3,772	4,186	-370	3,436	3,886	-729	3,037	3,557
		0.200	> 109	2,671	23	3,318	3,769	-707	2,830	3,359	-972	2,340	2,859

¹ For first encounters in data-generating models ${\mathcal{M}}_b$, ${\mathcal{M}}_{bh}$ and ${\mathcal{M}}_{tbh}$.

LLM-AIC denotes selection of the AIC-best loglinear model, LLM-BMA denotes Bayesian model-averaging of loglinear models, and BLCM denotes nonparametric Bayesian latent-class model estimation.