Direct answers to the questions review teams actually ask.
The honest answer is two - but heterogeneity estimates, prediction intervals, and publication-bias tests each need more. What the methods literature actually says.
4 min read · 2026-08-29