A comparison of missing data procedures for addressing selection bias in HIV sentinel surveillance data

Table 4 Probability coverage of 95% confidence intervals for $β_{1}$ , $β_{1} = 0$

Level of selection bias	Complete-case (Fixed-effects)	Complete-case (Mixed-effects)	Model 1	Model 2	Model 3
High	0.038	0.884	0.059	0.035	0
Moderate	0.129	0.933	0.593	0.179	0.007
Low	0.12	0.924	0.603	0.685	0.032

ISSN: 1478-7954