[Purpose/Significance]The traditional citation analysis method based on statistical counting is criticized for not considering the motivation of the author, especially the long period prejudice on self-citation. It is generally believed that self-citation is an excessive rec-ognition and trust to his own previous work by the author and thus will mislead the development of science. [Method/Process]This paper attempts to study the evaluation of self-citation from the perspectives of citation polarity, citation position and citation density. We analyze the citation motivation and the citation background to evaluate whether there is prejudice in previous self-citation evaluation; we discuss ci-tation level, citation position and the potential adjustment of the association between contexts to the same reference. The study uses qualita-tive and quantitative methods to analyze academic publications of different subjects and disciplines. [Results/Conclusion]The statistical a-nalysis shows that there is no deviation in the reference density and polarity uncertainty, and there is a bias in the enthusiasm and negativity of the citation. The effect of self-cited behavior on citation density, location concentration, and some reference patterns associated with it is revealed.%目的/意义 建立在统计计数基础上的传统引文分析方法,由于不考虑施引作者的动机而受到批判,尤其是对自引一直存在偏见,认为自引是作者对自己作品的过度肯定,会误导科学发展.[方法/过程 通过引用极性、引用位置和引用密度三个角度来讨论对自引的评价问题,通过探析引用发生的动机和引文背景来研究以往对自引的评价是否存在偏见;探讨引文水平和引文位置与上下文之间的关联对相同参考文献的潜在调节作用.采用定性和定量方法分析了不同主题和学科的学术出版物的自引情况.[结果/结论 统计分析结果表明自引在引用密度和极性不确定性方面没有偏差,在引用的积极性和消极性方面存在偏差;揭示了自引行为对引文密度、位置集中和关联的一些引用模式的影响.
展开▼
