. 2019 Apr 26;14(4):e0215652. doi: 10.1371/journal.pone.0215652

Table 2. Estimated variance components, intraclass correlations and model fit statistics for different specifications of the multilevel models for nonresponse to self-administered questionnaire, Belgian Health Interview Survey, 2013.

Model	0	1	2	3
	Intercept only	With respondent characteristics	With interview circumstance	With interviewer characteristics
Variance components
Not scaled:
$σ_{e}^{2} (i n d i v i d u a l l e v e l)$	3.290
$σ_{u}^{2} (i n t e r v i e w e r l e v e l)$	2.869	2.663	2.879	2.593
Scaled
$σ_{e}^{2} (i n d i v i d u a l l e v e l)$		3.072	3.025	2.898
$σ_{u}^{2} (i n t e r v i e w e r l e v e l)$		2.486	2.654	2.284
ICC _interviewers	0.466	0.447	0.467	0.441
PCV		13.3%	7.5%	20.4%
Model fit
Log likelihood	-3099.214	-2442.017	-2416.269	-2410.129
LR-Test vs. logistic model (p-value of LR test)	1469.44 (0.000)	802.09 (0.000)	807.92 (0.000)	760.09 (0.000)
LR-Test against previous model (df; p-value of LR test)		1314.394 (7; 0.000)	51.496 (2; 0.000)	12.280 (1; <0.005)