Optimizing the maximum reported cluster size for the multinomial-based spatial scan statistic

Moon, Jisu; Kim, Minseok; Jung, Inkyung

doi:10.1186/s12942-023-00353-4

International Journal of Health Geographics

Table 5 Multinomial model: simulation results for the true cluster model (B) and alternative hypothesis (4) using elliptical windows

From: Optimizing the maximum reported cluster size for the multinomial-based spatial scan statistic

		Maximum reported cluster size (MRCS)																		Default setting
		1%	2%	3%	4%	5%	6%	8%	10%	12%	15%	20%	25%	30%	35%	40%	45%	50%	Overall	Default setting
SCIC₁	Freq^a	0	6	8	21	32	69	428	111	72	47	28	12	7	6	1	1	1	850	850
	Sen^b	NA	0.167	0.325	0.381	0.519	0.661	0.884	0.901	0.942	0.962	0.914	0.800	0.886	0.633	0.800	1.000	1.000	0.839	0.860
	PPV^c	NA	0.833	0.813	0.921	0.863	0.938	0.950	0.750	0.636	0.525	0.380	0.256	0.198	0.139	0.160	0.161	0.082	0.823	0.733
	Mis^d	NA	0.063	0.054	0.048	0.042	0.028	0.013	0.030	0.046	0.070	0.119	0.188	0.273	0.324	0.319	0.377	0.812	0.037	0.068
SCIC₂	Freq	0	7	8	26	35	69	401	101	65	44	27	21	16	11	5	6	8	850	850
	Sen	NA	0.171	0.350	0.446	0.549	0.658	0.884	0.895	0.935	0.959	0.911	0.819	0.900	0.727	0.800	0.900	1.000	0.835	0.860
	PPV	NA	0.857	0.813	0.913	0.864	0.938	0.951	0.744	0.634	0.513	0.363	0.228	0.191	0.146	0.130	0.131	0.114	0.792	0.733
	Mis	NA	0.062	0.053	0.045	0.040	0.028	0.012	0.032	0.046	0.074	0.135	0.222	0.289	0.331	0.406	0.442	0.576	0.054	0.068
Elbow	Freq	0	5	9	35	51	74	423	99	65	39	18	16	10	4	1	1	0	850	850
	Sen	NA	0.160	0.422	0.560	0.608	0.654	0.887	0.905	0.935	0.949	0.911	0.800	0.800	0.750	0.800	1.000	NA	0.833	0.860
	PPV	NA	0.800	0.833	0.914	0.869	0.932	0.947	0.755	0.638	0.493	0.320	0.216	0.164	0.173	0.160	0.161	NA	0.829	0.733
	Mis	NA	0.064	0.047	0.037	0.036	0.029	0.013	0.029	0.045	0.081	0.167	0.232	0.310	0.283	0.319	0.377	NA	0.038	0.068
MCS-P	Freq	0	4	6	25	41	66	380	106	86	56	39	13	12	9	4	3	0	850	850
	Sen	NA	0.100	0.333	0.568	0.600	0.630	0.878	0.900	0.944	0.971	0.949	0.923	0.800	0.867	0.950	1.000	NA	0.848	0.860
	PPV	NA	0.500	0.833	0.921	0.878	0.908	0.952	0.761	0.648	0.545	0.424	0.308	0.220	0.194	0.184	0.163	NA	0.801	0.733
	Mis	NA	0.072	0.053	0.035	0.036	0.032	0.013	0.028	0.042	0.062	0.099	0.156	0.220	0.272	0.308	0.372	NA	0.039	0.068
MCHS-P	Freq	0	3	5	21	41	64	372	106	86	59	41	18	15	12	4	3	0	850	850
	Sen	NA	0.133	0.320	0.562	0.610	0.644	0.882	0.900	0.942	0.966	0.946	0.889	0.827	0.817	0.950	1.000	NA	0.853	0.860
	PPV	NA	0.667	0.800	0.915	0.882	0.926	0.950	0.758	0.644	0.529	0.409	0.276	0.213	0.178	0.184	0.163	NA	0.787	0.733
	Mis	NA	0.068	0.055	0.036	0.035	0.030	0.013	0.029	0.043	0.068	0.107	0.187	0.236	0.292	0.308	0.372	NA	0.044	0.068

^aFreq: frequency
^bSen: sensitivity
^cPPV: positive predictive value
^dMis: misclassification
Sensitivity, PPV, and misclassification rate at the most frequently selected optimal MRCS are shown in bold

Back to article page

ISSN: 1476-072X

Contact us

Submission enquiries: journalsubmissions@springernature.com