网络环境下国内学者引证行为变化与学科间差异——基于历时角度的分析

Differences of Citing Behavior over Time and Across Fields in China: A Diachronous Analysis

引证是学者最基本的学术行为之一,新时期引证行为的变化与趋势引起了图书情报学者及各学科的关注。本文从文献计量学的历时分析角度,基于学科层面,分析和比较我国学者引证行为的变化特征。选择"中国引文数据库"中四个学科(哲学、图书情报、物理和机械工程)1994—2013年内896 645篇论文的引文数据进行分析。结果显示:1总体上不同时段、不同学科间的引证情况存在显著差异,各年发表论文被引量曲线呈拱形,被引的累积效应不明显;2发表在不同年份的论文,特定年段后(时间窗),论文平均被引量呈现一致变化(在早期快速增加,近年来细微下降),未被引论文比例在早期下降明显,近年来趋于稳定或细微增长;3发表在特定年份的论文,其平均被引峰值,机械工程和哲学分别为7年和9年,物理学和图书情报学都为3年。图书情报学中各年论文被引峰值时间较为稳定;另外三个学科则在早期快速增加,近期趋于稳定;近年来,四个学科显现相对一致的被引特征,国内学者引证行为向相对稳定状态发展。图9。参考文献32。

Citing is a fundamental academic behavior among scholars; the distribution and change of citing behavior has been studied extensively by information scientists. In this paper,we analyse and compare the difference of the received citations in discipline level over a period of time and across fields in China by implementing the diachronous methods of bibliometrics. Citations of 896 645 papers from the Chinese Citation Database（ 1994 to 2013） and from four disciplines,namely,Philosophy,Library and Information Science（ LIS）,Physics,and Mechanical Engineering,are collected. We focus on the following： 1） the general differences of citation distributions among disciplines,2） the citation or uncitedness characteristic of papers published in different years（ for example,papers published in 2000,2001,2002... are cited respectively after 5 years,that is,2004,2005,2006...）,and 3） the citation characteristic of papers cited in different years（ for example,a paper published in 2000 is cited in 2000,2001,2002...）.Results indicate the following conclusions. 1） The growth of published papers is generally S-shaped and undergoes three stages（ i. e.,slow,rapid,and slow growth）. The curves of the received citations in four disciplines exhibit an arch shape（ i. e.,the middle is high and the end is low）. The cumulative phenomenon of received citations is not obvious. The received citations strongly differ across different fields and over time. 2） About the citation and uncitedness characteristic of papers published in different years. Firstly,the average of the received citations in a given year window changes identically,initially increases rapidly,and then slightly decreases in the recent years. The average of the received citations exhibits significant differences among the four disciplines in different time spans. In one-,two-,and three-year time window,a maximum value is observed in LIS in each published year. The value slowly decreases and the LIS obtains a minimum value within the 10-year windows. However,Physics and Mechanical Engineering exhibit a totally opposite change. Secondly,the uncitedness rate generally decreases in the early stages of the study period,but stabilises or even increases slightly in recent years. The uncitedness rates in the one-year window are relatively stable,but decrease rapidly in the two-year window and drop sharply in the five-year window. 3）About the citation characteristic of papers cited in different years. Notable differences exist among the trends of the mean of the received citations of the different fields. The maximum values of the average of the received citations peak after seven years for Mechanical Engineering,nine years for Philosophy,and three years for both Physics and LIS. Thus,those higher values of citations for Mechanical Engineering and Philosophy are not captured in calculating Two-year Journal Impact Factors. Moreover, present findings also suggest that results of the received citations obtained for a specific field should not be generalized at the micro level,nor used to describe exactly the situation of other or larger fields. About the citation characteristics of papers cited in different years,LIS citations are completely different from those of the other three disciplines. LIS citations peak at two or three years and then slightly decrease in all cited years. The curves of the other three disciplines are similar. Papers published in the early stages of the study period have a later cited peak. Recently,all four disciplines indicate relatively consistent citation trends and the citation behavior of Chinese scholars has been relatively steady. 9 figs. 32 refs.