Single paper meta‐analysis is unavoidable
主张行为研究论文中应联合分析多个实验，用统计汇总代替非统计汇总，并比较了单篇元分析与传统元分析的优劣，通过案例展示其好处。
Abstract We advocate that the multiple studies of a common phenomenon that are featured in a typical behavioral research paper be jointly analyzed to provide a statistical summary of the set of studies as a whole. Indeed, we view such single paper meta‐analysis as unavoidable in typical behavioral research papers because (i) such papers feature multiple studies of a common phenomenon, (ii) such papers contain summaries of those studies, and (iii) statistical summaries of studies (i.e., meta‐analyses) are superior to non‐statistical summaries of them. Nonetheless, the current dominant practice is for such papers to feature separate statistical analyses of the data from each of the studies but to contain a non‐statistical summary of the multiple studies rather than a statistical summary. We believe that this is regrettable and therefore aim to rectify matters. Consequently, we review some considerations about meta‐analysis and statistical analysis more broadly; compare single paper meta‐analysis to traditional meta‐analysis and illustrate its benefits via a case study; discuss and dismiss concern about single paper meta‐analysis; and discuss single paper meta‐analysis and the review process.