基于DEA-Malmquist的学术期刊生产率分析

Analysis of academic journals productivity based on DEA-Malmquist

【目的】研究学术期刊的生产率及其变化问题。【方法】通过数据包络分析(DEA)模型评价学术期刊的表现,讨论出版期数、引文密度、被引半衰期等指标的适用性,利用Malmquist指数测算期刊表现的变化情况。【结果】2007—2013年,中国社会科学引文索引(CSSCI)管理学领域期刊表现年均进步幅度为5.5%。进入国家自然科学基金委员会认定期刊列表的期刊的进步程度比未进入列表的期刊要快,各期刊表现的差异程度也相对较小,"马太效应"显现。进入或退出CSSCI名单的期刊的表现一般在第2年发生较大变化,表明存在"牛鞭效应"。【结论】DEA-Malmquist方法适用于多指标评价期刊生产率及其表现变化。由于引用产生过程的"蜂窝煤"特性,与BCC模型相比,CCR模型更适合评价期刊表现。

[Purposes] This paper aims to analyze the productivities of academic journals and their changes. [Methods] The performance of academic journals was evaluated by data envelopment analysis model, the applicability of indexes such as annual publications, citation density and cited half-life was discussed, and the changes of performance of journals were calculated based on Malmquist index.[ Findings] The annual average performance progress of journals in management science field indexed in China Social Science Citation Index （CSSCI） was 5.5% from 2007 to 2013. The progress of the listed journals identified by National Nature Science Foundation of China was greater than that of the non-listed journals, and the degree of difference in the performance of journals was relatively small, reflecting the existence of the Matthew effect. The performance of journals changed greatly two years after entering or exiting CSSCI, indicating the presence of the Bullwhip effect. [ Conclusions ] DEA-Malmquist method is suitable for dealing with multi-index evaluation of productivity of journals and their changes. The CCR model is more suitable for evaluating the performance of journals than the BCC model due to the honeycomb briquette characteristics of the reference generation process.