Skip to main content
Fig. 3 | Malaria Journal

Fig. 3

From: Bias in logistic regression due to imperfect diagnostic test results and practical correction approaches

Fig. 3

Results from the standard logistic regression (black) and Bayesian model 3 (blue). Left panel shows that inference regarding disease risk factors can be substantially different when using the standard logistic regression and Bayesian model 3. Stars indicate 95 % confidence/credible intervals that did not include zero. The ‘Time’ covariate refers to time of residence in the region. Right panels show that both microscopy sensitivity and infection probability decrease as a function of time living in this region

Back to article page