PDF(Sci-hub)
Abstract：
There is extensive choice in R to support meta-analyses.Two packages in this ecosystem include meta and metafor and provide an excellent opportunity to apply a structured checklist previously developed for contrasts between R packages relevant to challenges in ecology and evolution.Meta is a direct, intuitive choice for rapid implementation of general meta-analytical statistics. Metafor is a comprehensive package best suited for relatively more complex models.Both packages provide estimates of heterogeneity, excellent visualization tools, and functions to explore publication bias.The package metafor has a steeper learning curve but greater rewards. Reference to the learning curve and capacities of the statistical software Stata provided a benchmark outside the R ecosystem and confirmed the consistency in statistics.The usefulness of meta-analyses is not just in the synthesis of the research but in the process of doing the scientific synthesis. Reporting of contrasts and checks for robust statistics is an important contribution to more transparent and reproducible scientific syntheses.
摘要：
R中有广泛的选择来支持荟萃分析。这个生态系统中的两个包包括meta和metafor，并提供了一个很好的机会来应用之前开发的结构化清单，用于与生态和进化中的挑战相关的R包之间的对比。Meta是直接的，快速实施一般荟萃分析统计的直观选择。Metafor是一个全面的软件包，最适合相对更复杂的模型。这两个软件包都提供了对异质性的估计，优秀的可视化工具，以及探索出版偏见的功能。包元有一个更陡峭的学习曲线，但更大的回报。对统计软件Stata的学习曲线和能力的引用提供了R生态系统之外的基准，并证实了统计的一致性。荟萃分析的有用性不仅在于研究的综合，还在于进行科学综合的过程。报告对比和检查可靠的统计数据是对更透明和可重复的科学综合的重要贡献。
