中国科技期刊研究 ›› 2019, Vol. 30 ›› Issue (9): 937-943. doi: 10.11946/cjstp.201903050140

• 论坛 • 上一篇    下一篇

被撤销论文的学术不端行为及应对策略研究

田瑞强1,)2),姚长青2),潘云涛2)   

  1. 1)武汉大学信息管理学院,湖北省武汉市武昌区八一路299号 430072
    2)中国科学技术信息研究所,北京市海淀区复兴路15号 100038
  • 收稿日期:2019-03-05 修回日期:2019-06-18 出版日期:2019-09-15 发布日期:2019-09-15
  • 作者简介:田瑞强(ORCID:0000-0002-7399-7193),硕士,助理研究员,E-mail: tianrq2011@istic.ac.cn|姚长青,博士,研究员|潘云涛,硕士,研究员。

Academic misconduct in retracted articles and solution strategies

TIAN Ruiqiang1,)2),YAO Changqing2),PAN Yuntao2)   

  1. 1)School of Information Management, Wuhan University, 299 Bayi Road, Wuchang District, Wuhan 430072, China
    2)Institute of Science and Technology Information of China, 15 Fuxing Road, Haidian District, Beijing 100038, China
  • Received:2019-03-05 Revised:2019-06-18 Online:2019-09-15 Published:2019-09-15

摘要:

【目的】 分析被撤销论文中包含的学术不端行为,研究应对策略,为科研诚信建设提供参考建议。【方法】 基于Retraction Watch数据库统计了自1955年以来全球范围的被撤销论文,分析论文被撤销的主要原因,并从科研诚信教育、治理机制、制度建设和公开透明原则等方面讨论相应的应对策略。【结果】 重复、错误、不可再现、造假是论文被撤销的主要原因,因其被撤销论文数量约占被撤销论文总量的70%;因数据问题被撤销的论文占比为23%。中国论文的被撤销原因以错误和重复为主,美国则以错误和不可再现为主。被撤销论文的平均撤销时滞为1095天,造假论文因更不容易被发现,撤销时滞最长,达到1598天。基于职业成长的科研诚信教育、基于中心化模型的多方合作治理、学术不端治理的制度建设及公开透明的学术出版等方面的共同建设,有望有效抑制学术不端行为。【结论】 对于重复这类易识别的学术不端行为,通过采取相应的技术和控制流程有望有效防控;对于不可再现、造假等不易识别的学术不端行为, 则需要从科研诚信教育、多主体的合作治理和长效体制机制建设方面入手。此外,促进科学数据的公开、透明对于抑制与数据问题相关的学术不端非常必要且有效。

关键词: 学术不端, 学术不端治理, 科研诚信, 被撤销论文

Abstract:

[Purposes] This paper analyzes the academic misconduct contained in the retracted articles and studies the solution strategies, in order to provide reference for the construction of scientific research integrity. [Methods] Based on the Retraction Watch database, we collected the global retraction articles from 1955 to 2017, analyzed the main reasons for the articles being retracted, and discussed the solution strategies from the aspects of scientific research integrity education, governance mechanism, institution construction, and openness and transparency. [Findings] Duplication, error, irreproducibility, and falsification are the main reasons for retracting articles, accounting for about 70% of the total number of retracted articles; 23% of the retracted articles are involved with data issues. The reasons for the revocation of Chinese papers are mainly errors and repetitions, while the United States is dominated by mistakes and irreproducibility. The falsification paper is less likely to be discovered, and the time lag is the longest, reaching 1598 days. The joint construction of scientific research integrity education based on career growth, multi-party cooperative governance based on centralized model, institutional construction of academic misconduct governance, and open, transparent academic publishing is expected to effectively deal with academic misconduct. [Conclusions] It is expected to effectively reduce the duplication of such identifiable academic misconduct by adopting appropriate technical and control processes. For the irreproducible and falsification academic misconduct, we need to start with scientific research integrity education, multi-agent cooperative governance, and long-term institutional mechanism construction. In addition, promoting openness and transparency of scientific data is necessary and effective for the suppression of academic misconduct related to data issues.

Key words: Academic misconduct, Governance of academic misconduct, Research integrity, Retracted article