Exploring the importance of controlling heteroskedasticity and heterogeneity in health valuation: a case study on Dutch EQ-5D-5L

Table 4 Grade-of-membership (GOM) of the scale-adjusted latent class (SALC)

	GOM for taste class 2 garbage class (29% of respondents)		GOM for scale class 2 more random class (59% of respondents)
	Coef	p-value	Coef	p-value
Intercept	0.5022	0.098	0.5453	0.117
Female	0.5173	0.022	1.0399	0.882
Age in years		< 0.001		0.875
16–35	2.4143	0.005	1.1669	0.554
36–54
above 55	0.2100	0.011	1.4360	0.141
Educational attainment*		0.686		0.697
Low	0.6366	0.388	0.8678	0.705
Medium
High	0.9140	0.839	0.9635	0.911
Chronic Illness
Yes	1.0643	0.895
VAS score Health
Below 70	1.8346	0.249
70 >	–
Difficulty level
Failed dominant task			3.1286	0.044
Found tasks easy			1.1337	0.834
Found tasks hard			0.6152	0.140

Results are shown on the odds ratio scale. For education, the lowest group included up to the primary, the medium group included secondary to associates, and the highest group included bachelor's degrees and above. The standard errors are shown in Appendix 1 Table 8

ISSN: 1477-7525