医学论文“统计学处理”常见问题分析及建议

Analysis on common problems of “statistical analysis”in medical papers and suggestions

【目的】分析医学论文"统计学处理"中存在的问题,为提高医学期刊学术及编辑质量提供参考。【方法】审读已发表的医学论文,结合典型案例,对其中的"统计学处理"问题进行归纳总结。【结果】医学论文"统计学处理"部分主要存在下列问题:"统计学处理"内容不全,多组比较未交待两两比较的方法,统计软件与统计推断方法混淆,统计学基本概念模糊,统计描述与统计推断方法不一致,"统计学处理"与"结果"内容不一致,"统计学处理"中含有"结果"部分的内容。【结论】为了避免"统计学处理"中的错误,编辑应加强对统计学知识的学习,注意全文各部分统计学内容之间的联系;期刊应建立统一的编辑规范,严格按照正确的统计学方法修改论文;审稿中应充分发挥统计学专家的作用。

[Purposes]This paper aims to analyze the problems of ＂statistical analysis＂in medical papers and provide references for improving academic and editorial quality. [Methods] Many published medical papers were reviewed and the problems about＂statistical analysis＂ were summarized from typical cases. [Findings] The following problems are common in ＂statistical analysis＂： 1 the content of ＂statistical analysis ＂ is incomplete; 2 the multiple comparison method is not introduced in the comparison of three or more samples; 3 the statistical software and the statistical inference method are confused; 4 the basic statistical concepts are misunderstood; 5 the statistical description is inconsistent with the statistical inference; 6 the content in＂statistical analysis＂is inconsistent with the＂result＂content; 7and the＂statistical analysis＂includes the content which should be included in ＂results＂. [Conclusions] To avoid problems in ＂statistical analysis＂,editors should strengthen statistical learning and pay attention to the consistency of statistical contents in the whole paper; uniform editorial standards should be set up,and the papers should be modified according to right statistical methods; statistical experts should be invited to reviewpapers including complex statistical contents.