文章摘要
利用统计方法与规律发现论文数据造假
To expose data fabrication in papers by statistical discipline and methods
投稿时间:2018-04-13  修订日期:2018-05-02
DOI:
中文关键词: 学术不端  论文造假  数据造假  统计学  数据审查
英文关键词: Academic misconduct  paper fabrication  data fabrication  statistics  data check
基金项目:
作者单位E-mail
刘清海 中山大学 学报编辑部 liuqh@mail.sysu.edu.cn 
摘要点击次数: 32
全文下载次数: 
中文摘要:
      期刊编辑在防范学术不端过程中起着重要作用,其一即发现问题。期刊编辑可以在EXCEL中验证统计学结果,以发现统计数据造假;可以利用统计检验量分布值与P值对应规律以及区间值与P值对应规律,发现数据造假。这些方法为防范学术不端体系增加了利器,期待有更多学者发掘相关规律与分享成果。如欲发现论文的数据造假,期刊编辑首先须提高数据审查意识,其次应提升统计学素养与能力。发现数据造假,还必须要求来稿提供统计检验量值与精确P值。在论文评议流程中,建议增设数据与统计学审查流程,以加强对数据的把关力度。
英文摘要:
      Journal editors play an important role in preventing academic misconduct. To detect statistical fabrication, journal editors can verify statistical results in EXCEL, and use the statistical law of the tested distribution value and corresponding P value or the law of interval value and corresponding P value. These findings are helpful to prevent academic misconduct, and it is expected that more scholars join the group to explore the relevant laws and share the findings. To find out the data falsification, journal editors must firstly improve their awareness of data review, and secondly, enhance their statistical literacy and ability. In addition, the tested distribution value and accurate P value are needed in papers. It is suggested that data and statistical review process be inserted in paper review process so as to strengthen the data check.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭