Skip to main content

Table 6 Bias in β ^ 1 and β ^ 3 for scenarios with differential trends in intervention and non-intervention sites (with a high level of selection bias)

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

True parameters

Complete-case (Fixed-effects)

Complete-case (Mixed-effects)

Model 1

Model 2

Model 3

Scenario 1

     

β 1 = 0

−0.082

−0.002

−0.182

−1.983

−0.438

β 3 = –0.025

−0.037

−0.024

−0.092

−1.388

−0.362

Scenario 2

     

β 1 = –0.025

−0.085

−0.002

−0.181

−1.764

−0.415

β 3 = –0.025

−0.006

0.001

−0.066

−1.188

−0.307