Skip to main content

Table 2 Cohen’s \(\kappa\) statistics for assessing the normality based on the Q–Q plots

From: The normality assumption on between-study random effects was questionable in a considerable number of Cochrane meta-analyses

 

Binary outcome

(N = 2498a)

Non-binary outcome

(N = 2533b)

 

Originalc

ORd

RRd

RDd

Originala

Cohen’s \(\kappa\) statistic (I)e

0.36

0.35

0.35

0.36

0.37

Cohen’s \(\kappa\) statistic (II)f

0.44

0.42

0.43

0.44

0.46

  1. aBased on 2498 eligible meta-analyses with P-values ≥ 0.1 (from the SW test) and the between-study variance estimates > 0 for all of the OR, RR, and RD for binary outcomes
  2. bBased on 2533 eligible meta-analyses with P-values ≥ 0.1 (from the SW test) and the between-study variance estimates > 0 for non-binary outcomes
  3. cOriginally reported effect measures in Cochrane meta-analyses
  4. dOdds ratios, relative risks, and risk differences re-calculated from 2 \(\times\) 2 tables for meta-analyses with binary outcomes
  5. eBased on all 75 categories for the 5 scores for the left tail, 5 scores for the right tail, and 3 scores for the multimodal status
  6. fBased on 2 aggregate categories of normality (all scores equal 0) vs. non-normality (any score does not equal 0)
Back to article page

BMC Medicine

ISSN: 1741-7015

Contact us