非参数统计的期刊影响力评价方法研究

首先通过对期刊以及机构评价中的标准化指标进行较深入的比较,分析在学术期刊评价中不同类型指标对评价结果有效性的理解,探寻学术期刊影响力的主要属性特征及等效测度指标。其后按照普赖斯提出的知识累积金字塔式模式,构建基于非参数统计的评价指标-分布权秩得分(分布得分),该指标按照对数法则对不同引用量区间的文献赋予不同权重,并考虑研究领域引文网络的节点中心度,选择入度中心度(被引量)衡量文献在知识传播中的位置。最后,选择核心版SCI生物学领域的46种期刊作实证分析,并将本文所构建的分布得分指标与常用的期刊评价指标做对比分析。

Journal Influence Evaluation Methods Based on Nonparametric Statistics

By deeply comparing the standardized indicators of journal and institution evaluations, this paper analyzes the comprehensibility difference to the validity of the evaluation results, to explore the influence property of academic journals and equivalent measurement indicators. According with Gold tower mode of knowledge accumulation proposed by Derek de Solla Price, we construct an indicator based on nonparametric statistics. This indicator gives different weights to the different documents with different citations range according to logarithmic law, considers the node centrality of the research field citation network, and selects the in-degree centrality-the citations to measure the location of the papers in the dissemination of knowledge. Finally, it chooses 46 journals in the field of biology according with the core version of the SCI as the evaluation objects, and makes a comparative analysis on the distribution score index with other popular journal evaluation indicators.