The recent discussion about the non-reproducibility of scientific research results¹ brought statistical methods used in laboratory animal studies under closer inspection (see, e.g., the recent book by Lazic² for an extensive discussion of the wide range of problems involved). Minimal standards were defined that should be applied in preclinical studies involving animals,^3,4 but in practice, these existing guidelines are often not followed. The warnings of statisticians about the consequences of the commonly applied procedures of study design and data analysis have a long history,^5,6 but it seems only recently that the research community has become widely aware that there actually is a severe problem and that action has to be taken to improve the quality of preclinical research studies.^7,8

中文翻译：

---

动物试验中的实验重复

最近关于科学研究结果的不可重复性的讨论¹使实验室动物研究中使用的统计方法受到更仔细的检查（例如，参见 Lazic ²最近的书，对所涉及的范围广泛的问题进行了广泛的讨论）。定义了应在涉及动物的临床前研究中应用的最低标准，^3,4但在实践中，这些现有指南往往不被遵守。统计学家关于研究设计和数据分析的常用程序后果的警告由来已久，^5,6但似乎直到最近，研究界才广泛意识到实际上存在一个严重的问题，必须采取行动来提高临床前研究的质量。^7,8