摘要：

【目的】 通过案例分析，探讨科技期刊应对研究论文可重复危机的路径，促进科技期刊更好地发挥学术守门人的作用。【方法】 通过网络调研法、文献调研法深入分析癌症生物学论文再现性项目（Reproducibility Project：Cancer Biology，RP：CB）的实施过程、实验结果、实验可重复率低的原因，深入探究RP：CB项目的过程与结果。【结果】 RP：CB项目的可复制性数据一共产生158个效应，包括正向效应和无效效应。对于以数值形式报告的原始正效应，重复实验的中位效应量仅为0.43，比原始效应量的中位值2.96小85%，这说明复制实验的结果远达不到原始研究结果的效应量。【结论】 科技期刊可以从使用新一代期刊评价体系——TOP Factor、增加注册报告发文比例及呼吁科研人员重视研究的可重复性等方面推动科学研究向着透明、可重复的方向发展。

关键词: 可重复危机, 发表偏倚, TOP Factor, 注册报告

Abstract:

[Purposes] This article aims to explore measures for scientific journals to address the crisis of paper reproducibility through case analysis, so as to help scientific journals better play the role of academic gatekeepers. [Methods] This article conducted an in-depth analysis of the implementation process, experimental results, and reasons for low reproducibility of the Reproducibility Project: Cancer Biology (RP: CB) through online research and literature review methods and explored the process and results of the RP: CB. [Findings] The replicability data of the RP: CB has generated a total of 158 effects, including positive and ineffective effects. For the original positive effects reported in numerical form, the median effect size of the repeated experiment is only 0.43, which is 85% lower than the median of the original effect size of 2.96. This indicates that the results of the repeated experiment are far from reaching the effect size of the original research results. [Conclusions] Scientific journals can promote the development of scientific research towards transparency and repeatability by using a new generation of journal evaluation system, namely TOP Factor, increasing the proportion of registered reports and publications, and calling on researchers to pay attention to the repeatability of research.

Key words: Repeatable crisis, Publication bias, TOP Factor, Registered report