Meta-evaluation of meta-analysis: ten appraisal questions for biologists.

Also, ES1–3 are all what people describe as ‘highly significant’, but their magnitudes of effect, and thus biological relevance, are very different.

also in 132,142 other papers

However, if presented without the associated τ 2 , Q can be misleading because, as is the case with most statistical tests, Q is more likely to be significant when more studies are included even if τ 2 is relatively small (see also Q7); the reader should therefore check whether both statistics are presented.

also in 1,824 other papers