摘要：

【目的】 研究科技论文图表数据中的学术不端基本类型和表现特征,探讨期刊编辑部的具体识别和防范措施。【方法】 以PubPeer网站和相关文献中的真实事件为调研样例,总结图表数据学术不端的类型和表现特征,针对以上类型提出对应的甄别和防范措施。【结果】 图表数据学术不端行为主要分为图表数据伪造、剽窃和篡改三种类型。编辑可通过重视学术不端检测系统的结果分析、通过多种技术手段分析图片以及利用统计学方法检查表格数据来甄别图表数据学术不端行为。【结论】 期刊编辑部应加强以下措施来防范图表数据学术不端:重视学术诚信教育,在编辑加工环节制定详细的图片处理声明,要求作者提供原始数据,核查不同修改稿中的数据更改。此外,期刊编辑部还应建立完备的学术不端处理机制并积极推动同领域期刊共享诚信档案,共同推进和谐学术生态的建设。

关键词: 学术不端, 图表数据, 数据伪造, 数据剽窃, 数据篡改, 案例分析

Abstract:

[Purposes] This paper aims to study the basic types and characteristics of data academic misconduct in images and tables, and discuss the specific idenpngication and prevention measures for editorial offices of journals. [Methods] We summarized the basic types and characteristics of data academic misconduct in images and tables in real cases on the PubPeer website and other literature, and then proposed idenpngication and prevention means. [Findings] Data academic misconduct in images and tables can be classified into fabrication, plagiarism, and falsification. Editors can idenpngy them by emphasizing the academic misconduct detection results, analyzing images through a variety of technical means, and checking the data in tables with statistical methods. [Conclusions] Journals should step up efforts to prevent data academic misconduct in images and tables. Specifically, they should attach importance to authors' academic integrity education, release detailed image processing standards, ask for original data, and check data changes in different revised manuscripts. For journals themselves, they should establish a sound mechanism for handling academic misconduct and actively promote the sharing of credible files among journals in the same field, so as to foster a correctitude academic atmosphere.

Key words: Academic misconduct, Data in images and tables, Data fabrication, Data plagiarism, Data falsification, Case analysis