影响因子中的期刊自引成分分析

Journal Self-citation Analysis in Impact Factors

基于2年影响因子的定义,推导出了影响因子与期刊前2年自引年度分数之和的关系式,发现影响因子与期刊前2年自引年度分数之和呈正相关.期刊自引文献在不同年份的分布是随机的,期刊自引年度分数与时间因子非线性相关,其峰值可出现在2年影响因子的计算周期内,用期刊他引率与影响因子之积(定义为表观他引影响因子)无法合理消除期刊自引的干扰.抽样调查20种样刊的自引情况,发现表观影响因子排序与他引影响因子(完全排除自引后的影响因子)排序差别较大,统计结果表明上述理论推断是正确的.建议国内外的期刊引证报告中应补充期刊自引半衰期指标或期刊累计自引年度分数,客观反映期刊自引文献的时间分布,以提高对期刊评价的客观性和公正性.

Based on the definition of 2-years impact factors, a formula between impact factor and self-citation year distribution coefficient was deduced. Citation literature distribution in different years is random and the peak can occur in 2 year impact factor calculation period, so the apparent its-citation impact factor（IF · Yo ）could not eliminate reasonably the interference of self-citation. A sample survey of 20 kinds of publications shows that rank ordering of apparent its-citation impact factor does not accord with that of its-citation impact factor（without self-citation）, so the above theory is correct. It is proposed that the self-citation half-life index or citation distribution coefficient should be added in JCR at home and abroad in order to enhance objectivity and impartiality of periodical evaluation.