Reinforcing Reproducibility and Replicability: An Introduction
Reinforcing Reproducibility and Replicability: An Introduction
复制标题
增强再现性和可复制性：简介
DOI：
10.1162/99608f92.9ba2bd43
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Connolly, Marie
中科院分区：
文献类型：
--
作者：
Vilhuber, Lars;Schmutte, Ian;Michuda, Aleksandr;Connolly, Marie
科学出版的目的是传播可靠的研究成果，让其接受同行的审查。这项工作的关键是记录这些发现的出处。研究和随后的出版过程中的科学实践、同行评审以及传播实践和工具，所有这些相互作用，以（希望）能够对科学主张的准确性进行有意义的讨论。然而，虽然所有实践和工具都对最终输出有所贡献，但有些实践和工具比其他实践和工具更少被讨论，而且不同学科的看法、使用和接受程度存在多种差异。在这个特殊主题以及后续名为“增强再现性和可复制性”的专栏中，我们将探讨这些主题，专家提供者和专家用户将提供他们的意见。虽然我们将从经济学学科这一特殊主题开始，但该专栏的关注范围不会那么狭隘，随着时间的推移，会提供来自其他学科的背景和声音。无论人们是否真的相信存在“复制危机”（Fanelli，2018），近年来人们对研究的可靠性表示了一些怀疑。部分回应是，人们越来越重视支持改进来源文档的各种方法。在社会科学领域，这包括预注册（Nosek et al., 2018, 2019）、预分析计划（Banerjee et al., 2020; Olken, 2015）、注册报告（Chambers, 2014; Hardwicke & Ioannidis, 2018; Journal of Development Economics, 2019）、跨学科工作论文和预印本的可用性经济学、统计学和物理学 (Vilhuber, 2020)，以及围绕数据和代码可用性日益严格的期刊政策，包括复制包的主动审查和验证 (Christian et al., 2018, 2020; Editors, 2021; Vilhuber, 2019)。
The purpose of scientific publishing is the dissemination of robust research findings, exposing them to the scrutiny of peers. The key to this endeavor is documenting the provenance of those findings. Scientific practices during the course of research and subsequent publication, peer review, and dissemination practices and tools, all interact to (hopefully) enable a meaningful discourse about the veracity of scientific claims. However, while all practices and tools contribute to the final output, some are less often discussed than others, and perceptions, usage, and acceptance differ in myriad ways across disciplines. In this special theme, and in a subsequent column called “Reinforcing Reproducibility and Replicability,” we will explore these topics, with expert providers and expert users providing their input. While we will start within the economics discipline in this special theme, the column will not be as narrowly focused, providing context and voice from other disciplines over time.Whether or not one actually believes there is a “replication crisis”(Fanelli, 2018), some doubts have been expressed in recent years about the reliability of research. Partially in response, there has been an increased emphasis on various methods that support improved provenance documentation. In the social sciences, this includes preregistration (Nosek et al., 2018, 2019), pre-analysis plans (Banerjee et al., 2020; Olken, 2015), registered reports (Chambers, 2014; Hardwicke & Ioannidis, 2018; Journal of Development Economics, 2019), greater availability of working papers and preprints across disciplines other than economics, statistics, and physics (Vilhuber, 2020), and increasingly more stringent journal policies surrounding data and code availability, including active review and verification of replication packages (Christian et al., 2018, 2020; Editors, 2021; Vilhuber, 2019).