摘要
[目的/意义]在期刊评价中,影响力与时效性是非常重要的两个方面,应该协调发展,但目前的绝大多数多属性评价方法,都是取长补短式性质的弥补式评价,难以评价期刊协调发展,需要解决这个问题。[方法/过程]提出因子几何平均法(FGM),首先通过因子分析提取公共因子,然后标准化后进行几何平均。以JCR 2015数学期刊为例进行了说明,发现影响力与时效性较高的期刊评价得分较高,因子几何平均法具有较高的区分度。由于因子分析在提取公共因子时存在信息损失,也会导致评价数据存在信息损失,另外要求评价指标具有较高的相关性。[结果/结论]要重视非弥补式评价的协调发展评价方法研究;因子几何平均法是一种有效的协调发展评价方法,适合比较成熟领域的学术期刊协调发展评价;采用因子几何平均法评价要注意学科差异。
[Purposes/Significance]In journal evaluation,the influence and timeliness,which are most important aspects,should develop harmoniously,but most of the multiple attribute evaluation methods are redeeming evaluation,which means that scores of evaluation index complement each other to make up their deficiencies in the total scores of evaluation,it is hard to evaluate harmonious development of periodicals.[Method/Process]In order to solve this problem,this paper puts forward factor geometric method(FGM),firstly,the common factors are extracted by factor analysis,and then implemented geometric mean after standardization.It takes JCR 2015 mathematical journal as an example,and finds FGM having good discrimination.The journals which have higher impact and timeliness can get better score.[Result/Conclusion]This paper shows that we should attach importance to the evaluation method of coordinated development of non-compensation evaluation;Factor geometric method is an effective method for the evaluation of coordinated development,which is suitable for the evaluation of more mature fields and the evaluation of coordinated development of academic journals;The difference of academic disciplines need be noticed with implementing the factor geometric method.
作者
俞立平
琚春华
Yu Liping;Ju Chunhua(Zhejiang Gongshang University,School of statistics and mathematics,Hangzhou 310018;Zhejiang Gongshang University,School of Management and E-business,Hangzhou 310018)
出处
《情报杂志》
CSSCI
北大核心
2019年第1期194-199,170,共7页
Journal of Intelligence
基金
教育部人文社会科学研究规划基金项目"协同创新深度的影响机制与对策研究"(编号:17YJA630125)
浙江省软科学重点课题"基于DEA的网上技术市场运行效率评价研究(编号:2018C25030)
关键词
期刊评价
因子几何平均法
非弥补式评价
journal evaluation
factor geometric method
non redeeming evaluation