Skip to main content
. 2022 May 16;13(6):716–744. doi: 10.1002/jrsm.1565

FIGURE 2.

FIGURE 2

Point estimates and performance metrics across all methods for each simulation scenario with N = 200. The model standard error for the matching‐adjusted indirect comparison outlier in the poor overlap scenario has an inordinate influence on the variability ratio; removing it reduces the variability ratio to 0.980 (0.019). Note that the version of STC evaluated does not actually target a marginal effect. [Colour figure can be viewed at wileyonlinelibrary.com]