研究论文支撑数据审查:逻辑、方法与实施框架

Review of supporting data for research papers:Logic,methodology,and implementation framework

【目的】从保障研究可检验性、可重复性出发,探索论文支撑数据审查的逻辑、方法和初步实施框架,为建立可循证操作、可细粒化验证、可自动计算的支撑数据审查机制提供基础。【方法】根据科技界和科技期刊界关于科研可靠性的基本原则,在简要总结研究论文支撑数据的政策与实践的基础上,利用系统化分析方法和循证设计方法,体系化设计和建立支撑数据审查的原则、架构、指标和实施建议。【结果】设计基于可计算证据链进行审查的原则机制,提出数据可用性声明审查、支撑数据FAIR度审查、支撑数据与研究结论对齐度审查的三层审查结构,初步提出可规则化、可操作、可测度、可评价、可计算的论文支撑数据审查框架和初步指标体系,并提出迭代实施原则与建议。【结论】研究有助于提高研究论文支撑数据审查的体系化程度、透明度、可评价性和实施效率,从而提高研究论文评审质量。

[Purposes]From the perspective of ensuring the testability and reproducibility of research,this paper explores the logic,methodology,and preliminary implementation framework for the review of supporting data in research papers,so as to provide a foundation for establishing an evidence-based operational,fine-grained verifiable,and automatically computable review mechanism for supporting data.[Methods]Based on the basic principles of scientific research reliability in the field of science and scientific journals,we briefly summarized the policies and practices of supporting data in research papers.Then,we used systematic analysis and evidence-based design methods to systematically design and establish principles,architecture,indicators,and implementation suggestions for supporting data review.[Findings]The principle mechanism is built based on computable chain of evidence.A threestep review structure that includes checking data availability statements,the compliance of the supporting data with FAIR principles,and the alignment of the supporting data with the study conclusions is proposed.The study presents a working framework and a firstphase indicator system for reviewing supporting data of research papers that are guidelines-based,operational,measurable,evaluable,and computable.[Conclusions]The study helps improve the systematic,transparent,and evaluable review process of supporting data for research papers,thus enhancing implementation efficiency and promoting the quality of peer review.