 The article discusses how to calculate the required information size in a meta-analysis. It proposes a new measure called diversity, D2, as an adjustment factor for the required information size. This measure is derived from the relative variance reduction when the meta-analysis model changes from a random effect to a fixed-effects model. The authors compare this measure to the commonly used measure of inconsistency, I2. They demonstrate that D2 is always greater than or equal to I2, regardless of the chosen between trial variance estimator. Therefore, D2 is a more accurate measure of the required information size in any random-effects meta-analysis. This article was authored by Gludd Christian, Brock Jesper, Thorland Christian, and others.