特征因子与其他文献计量指标关系研究——基于面板数据的估计

The relationship between eigenfactor and other bibliometric indicators based on panel data

为了研究期刊特征因子与其他文献计量指标间的关系,克服简单相关系数和普通回归分析的缺陷,本文基于汤森路透JCR纯数学期刊数据,采用面板数据(Panel Data)研究特征因子分值(Eigenfactor Score)及论文影响分值(Article Influence Score)与其他文献计量指标的关系。研究发现:特征因子分值和论文影响分值互为高度相关。即年指标及影响因子与特征因子分值无关,但和论文影响分值正相关。被引半衰期及5年影响因子与特征因子分值负相关,与论文影响分值正相关。总被引频次及期刊论文数量与特征因子分值正相关,与论文影响分值负相关。研究认为,特征因子的推出有利于期刊重视提高学术质量而不是提高文献计量指标值,对文献计量指标的相关关系的研究方法要重新审视,对于其他学科特征因子与文献计量指标的关系有待进一步研究。

The eigenfactor and other bibliometric indicators are analyzed based on panel data collected from Thomson Reuters JCR. The results show that eigenfactor score is highly correlated with article influence score. Immediacy index and impact factor are not correlated with eigenfactor score, however are correlated with article influence factor. Cited half-life and five -year impact factor are negative correlated with eigenfactor score, but are positive correlated with article influence factor. Total cites and articles are positive correlated with eigenfactor score, but are negative correlated with article influence score. It is argued that eigenfactor is very important for improving academic journals＇ quality. The methods for studying performance indexes relations should be reexamined. The relations among performance indexes in other disciplines require further research.