 The paper proposes a new measure called diversity, d2, to estimate the relative variance reduction when changing from a random effects to a fixed effects model in a meta-analysis. d2 is calculated as the percentage of between trial variability that contributes to the sum of between trial variability and sampling error estimate considering the required information size. The paper shows that d2 is different from the commonly used measure of heterogeneity, inconsistency, i2, which may underestimate the required information size. The authors also demonstrate mathematically that d2 is always greater than or equal to i2 for all meta-analyses. They conclude that d2 is a better alternative to consider model variation in any random effects meta-analysis and can adjust the required information size. This article was authored by Gluck Christian, Brock Jesper, Fall Unchristian and others. We are article.tv, links in the description below.