Skip to main content
Fig. 8 | BMC Medical Research Methodology

Fig. 8

From: Treatment of missing data in Bayesian network structure learning: an application to linked biomedical and social survey data

Fig. 8

Distribution of the difference in means of recall of three pairwise comparisons among three methods when there are 1000 data points: MICE’s increase over doing nothing (red), SEM’s increase over nothing (blue), and SEM’s increase over MICE (green) A. MCAR data. B. MAR data. C. MNAR data. The y-axis represents the difference of the mean recall (averaged over the 100 simulations). The x-axis represents the number of variables from 2-20. Column panels represent missing proportions

Back to article page