English

Simulations for the Q statistic with constant and inverse variance weights for binary effect measures

Methodology 2023-04-11 v2

Abstract

Cochran's QQ statistic is routinely used for testing heterogeneity in meta-analysis. Its expected value (under an incorrect null distribution) is part of several popular estimators of the between-study variance, τ2\tau^2. Those applications generally do not account for the studies' use of estimated variances in the inverse-variance weights that define QQ (more explicitly, QIVQ_{IV}). Importantly, those weights make approximating the distribution of QIVQ_{IV} rather complicated. As an alternative, we are investigating a QQ statistic, QFQ_F, whose constant weights use only the studies' arm-level sample sizes. For log-odds-ratio, log-relative-risk, and risk difference as the measure of effect, these simulations study approximations to the distributions of QFQ_F and QIVQ_{IV}, as the basis for tests of heterogeneity. We present the results in 132 Figures, 153 pages in total.

Keywords

Cite

@article{arxiv.2206.08907,
  title  = {Simulations for the Q statistic with constant and inverse variance weights for binary effect measures},
  author = {Elena Kulinskaya and David C. Hoaglin},
  journal= {arXiv preprint arXiv:2206.08907},
  year   = {2023}
}

Comments

Version 2 includes new Appendix C on empirical power of Q for log-odds-ratio

R2 v1 2026-06-24T11:55:23.310Z