图书情报工作 ›› 2022, Vol. 66 ›› Issue (20): 148-161.DOI: 10.13266/j.issn.0252-3116.2022.20.016

• 北京大学信息管理系成立75周年学术专辑 • 上一篇    下一篇

守正能否创新?——基于我国图情领域论文新颖性和传统性的分析

梁兴堃   

  1. 北京大学信息管理系 北京 100871
  • 收稿日期:2022-08-01 修回日期:2022-09-21 出版日期:2022-10-20 发布日期:2022-11-17
  • 作者简介:梁兴堃,助理教授,博士,E-mail:lxk@pku.edu.cn。

Novelty, Conventionality, and Scientific Impact of Papers in Library and Information Science in China:Evidence from Papers in CSSCI (2000-2019)

Liang Xingkun   

  1. Department of Information Management, Peking University, Beijing 100871
  • Received:2022-08-01 Revised:2022-09-21 Online:2022-10-20 Published:2022-11-17

摘要: [目的/意义] 以我国图情领域为例,测量论文的新颖性和传统性并探究其对论文学术影响力的作用进而揭示学术创新的规律。[方法/过程] 采用基于马尔科夫链蒙特卡罗(Markov chain Monte Carlo,MCMC)的方法,对我国2000年至2019年20年间在中文社会科学引文索引(CSSCI)中收录的图书馆学情报学领域的70 207篇研究论文的新颖性、传统性进行测量,并分析论文新颖性和传统性对论文学科影响力的作用。[结果/结论] 结果显示,其他因素不变时,论文新颖性提高1个单位,论文成为高被引论文的优势比增加11%,而论文传统性提高1个单位,论文成为高被引论文的优势比增加33%。边际效应分析显示,同时具有较高的新颖性和传统性的论文较之于其他类型的论文具有更高的成为高被引论文的可能性。此外,随着时间推移,新颖性对论文成为高被引论文概率的影响逐渐削弱,而传统性的影响逐渐增强。同时,作者团队规模对于论文的新颖性存在显著影响,这种影响随着时间的推移而增强。这些发现凸显我国图情领域守正创新的特点,为理解我国图情领域的学术创新规律提供新的实证基础。同时,也提出一种不同于传统信息计量的基于贝叶斯统计的新方法。

关键词: 新颖性, 传统性, 学科影响力, 高被引论文, 论文被引量, 马尔科夫链蒙特卡洛方法(MCMC), 团队作者, 贝叶斯统计

Abstract: [Purpose/Significance] With papers in the field of library and information science in China, this study measures the novelty and conventionality of these papers and explores their effects on the scientific impacts of these papers, in order to reveal the law of scientific innovation in this field. [Method/Process] This paper adopted a variation of the Markov Chain Monte Carlo (MCMC) approach to measuring novelty and conventionality of 70 207 papers in library and information science (n=70 207) from the Chinese Social Science Citation Index (CSSCI) in two decades, specifically, from 2000 to 2019. With robust logistic regression, this paper examined the impacts of papers' novelty and conventionality on their papers' scientific impacts. [Result/Conclusion] The results show that, ceteris paribus, the novelty of a paper increases by 1 unit, the odds ratio of the paper becoming a highly cited paper increases by 11% (p<0.000 1), and the conventionality of a paper increases by 1 unit, and the odds ratio of the paper becoming a highly cited paper increases by 33% (p<0.000 1). The marginal effect analysis shows that papers with high novelty and conventionality are more likely to be highly cited papers than other types of papers. In addition, over time, the impact of novelty on the probability of a paper being highly cited gradually weakens, while the impact of conventionality gradually increases. Meanwhile, author team size has a significant effect on the novelty of the paper, and such a effect increases over time. These findings highlight the characteristics of scientific innovation in the field of library and information in China, and provide novel empirical evidence to understand the law of scientific innovation in the field of library and information in China. Last but not least, this paper, based on Bayesian statistics, also proposes an alternative method for informetrics.

Key words: novelty, conventionality, scientific impact, hit paper, paper citation, MCMC, team authors, Bayesian statistics

中图分类号: