Skip to main content
Figure 4 | BMC Medical Research Methodology

Figure 4

From: A random effects variance shift model for detecting and accommodating outliers in meta-analysis

Figure 4

RVSOM statistics plotted against study number for the magnesium study. (a) Variance shift estimates, ω ^ j 2 . (b) Treatment effect estimates μ ^ . (c) Random effect variance estimates τ ^ 2 . (d) Likelihood ratio test statistics for a RVSOM for study j, plotted against study number. 95th percentile of the empirical distribution under the null hypothesis shown for the first k order statistics for each test: k = 1 (dotted line), k = 2 (dashed line) and k = 3 (solid line).

Back to article page