Skip to main content

Table 7 Adjusted means for the ten treatments

From: Network-meta analysis made easy: detection of inconsistency using factorial analysis-of-variance models

Treatment Adjusted mean Letter grouping
rosi 0.212    c
piog 0.317   b c
metf 0.318   b c
migl 0.496   b c
acar 0.605   b c
benf 0.709 a   c
vild 0.746 a   c
sita 0.876 a   c
SUal 1.029 a b  
plac 1.446 a   
  1. Means for the diabetes example of Senn et al. [7] were computed from model (2), dropping the design × treatment interaction (G.T) and modelling heterogeneity (G.S.T) as random. Pairwise comparisons at a family-wise Type I error rate of 5% by Edwards-Berry [20] test. Means with a common letter are not significantly different. The letter display was obtained by the method of Piepho [21]. Treatments are sorted in ascending order of means for ease of interpretation.