Abstract:

[Purposes] This paper summarizes the methods for discovering papers with scientific misconducts, hoping to provide references for journals to manage scientific misconducts and enhancing research integrity. [Methods] According to the retraction statements of 1834 papers on Web of Science from universities in China which were retracted for scientific misconducts, we through scientometrics methods, content analysis, and text measurement, analyzed the ways to find the scientific misconducts, investigation process, implementers of the retraction, and reasons for the retraction. [Findings] Third-party reporting is the main way for the discovery of scientific misconducts, and retractions are usually initiated by the editors-in-chief or editors of the journal after investigation and verification by the journal's editorial boards or publishers. The main reasons for retraction are fake peer review and plagiarism, and the reasons tend to be more complex and diverse. [Conclusions] Journals and publishers should provide special reporting channels and protect the privacy information of whistleblowers. Moreover, they should improve the criteria and policies for the retraction and develop a unified writing norm for retraction statements. In addition, efforts should be intensified to improve the paper check and review mechanism, strengthen the check and punishment, and pay attention to frequent, unusual, and proactive retractions.

Key words: Retraction statement, Retracted paper, Scientific misconduct, Publishing ethics, Text analysis

摘要：

【目的】 聚焦科研不端撤销论文,深入探究科研不端撤销论文的发现机制,为期刊参与对科研不端论文的治理提供参考,为出版伦理与科研诚信体系建设提供借鉴。【方法】以Web of Science数据库中1834篇我国高校科研不端撤销论文的撤稿声明为数据来源,综合使用科学计量、内容分析、文本计量方法,分析科研不端撤销论文的发现途径、调查过程、撤稿主体、撤稿原因等。【结果】第三方举报是发现科研不端论文的最主要途径,科研不端论文通常经期刊编辑部或出版商调查核实,由期刊主编或编辑发起撤稿。撤稿原因主要为虚假同行评议和剽窃,并且呈现出复杂化和多样化的趋势。【结论】期刊和出版商应当建立专门举报渠道,注重保护举报人隐私信息;不断完善对科研不端论文的撤稿依据和撤稿政策;制订统一的撤稿声明撰写规范;完善论文检测和审核机制,加大论文查处力度;关注频繁的、异常的、主动的撤稿行为。

关键词: 撤稿声明, 撤销论文, 科研不端, 出版伦理, 文本分析