Table 2Simulation parameters for proportions

1Distribution of summary proportions across studiesBeta, uniformf (see Appendix for details)
2True summary proportion, π0.001, 0.002, 0.005, 0.01, 0.02, 0.05, 0.10, 0.20, 0.30, 0.40, 0.50
3Number of studies, K5, 15, 30
4Sample sizes, N*Vectors of sample sizes for the following choices:
All sample sizes small (5-50 patients per study)
All sample sizes medium (51-200 patients per study)
All sample sizes large (201-1000 patients per study)
Mixed sample sizes – approximately 50% small, 40% medium, and 10% large.
5Heterogeneity, τ2Three levels: zero, small, and large. To determine τ, the square root of the heterogeneity variance, true summary proportions were multiplied by 0.10 or 0.50 for small or large heterogeneity, respectively.
6Correction factor, c0, 0.001, 0.01, 0.10, 0.5, 1, 2

For parsimony, in the Results section we present in detail scenarios corresponding to the underlined choices. The index j for the scenario has been dropped.


The exact values for sample sizes used in the simulations are given in the Appendix.

Some meta-analysis methods require the use of correction factors. See Continuity Correction Factors for details. The correction factor is an analytic choice and not a simulation parameter; however it is listed here for parsimony.


Strictly speaking, a uniform is a special case of the beta distribution, i.e., Beta(1, 1). The Appendix provides details on the parameters of the modal beta distributions used in the simulations.

From: Methods

Cover of Simulation-Based Comparison of Methods for Meta-Analysis of Proportions and Rates
Simulation-Based Comparison of Methods for Meta-Analysis of Proportions and Rates [Internet].
Trikalinos TA, Trow P, Schmid CH.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.