Year: 2021
Type: article
Abstract: Meta-analysis is an influential evidence synthesis technique that summarizes a body of research.Though impactful, meta-analyses fundamentally depend on the literature being sufficiently large to generate meaningful conclusions.Power analysis plays an important role in determining the number of studies required to conduct a substantive meta-analysis.Despite this, power analysis is rarely conducted or reported in published meta-analyses.A significant barrier to the widespread implementation of power analysis is the lack of available and accessible software for calculating statistical power for meta-analysis.In this paper, I provide an introduction to power analysis and present a practical tutorial for calculating statistical power using the R package metapower.The main functionality includes computing statistical power for summary effect sizes, tests of homogeneity, categorical moderator analysis, and subgroup analysis.This software is free, easy-to-use, and can be integrated into a continuous work flow with other meta-analysis packages in R.
Author: Jason W. Griffin
Institutions: Pennsylvania State University, University of Würzburg
Language: English
Subfield: Statistics, Probability and Uncertainty
Field: Decision Sciences
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W3138345420