摘要
相对于区域、产业层面的研究,企业创新绩效的测度样本通常包含大量的零值数据,因而会导致通常的创新绩效测度过程出现无解的现象。以往的研究大多选择将这一类样本直接剔除,而这种方法明显会损失大量的样本,使得对企业创新绩效的研究并不充分。文章将企业创新绩效的测度样本划分为五种类型,基于改进的DEA模型,建立了一种适用于包含大量零值数据的企业创新绩效测度体系,并对北京市某产业集群的样本企业进行了实证分析。结果显示,所构建的测度体系可以使样本损失率由传统方法的95.19%大幅下降至13.08%,更加充分地保留了原始样本的信息。
Compared with regional and industry-level studies,measurement samples of enterprise innovation performance usually contain a large number of zero-value data,which leads to the phenomenon that there is no solution in the usual innovation performance measurement process.Most of the previous studies chose to directly eliminate this kind of samples,but this method will obviously lose a large number of samples,making the research on enterprise innovation performance insufficient.This paper divides the measurement samples of enterprise innovation performance into five types,and then,based on the improved DEA model,establishes a measurement system for enterprise innovation performance containing a large number of zero-value data,and finally conducts an empirical analysis on the sample enterprises of an industrial cluster in Beijing.The results show that the proposed measurement system can significantly reduce the sample loss rate from 95.19%to 13.08%,with the original sample information more fully retained.
作者
陈鸣
孙红军
杨颖梅
Chen Ming;Sun Hongjun;Yang Yingmei(School of Management and Economics,Beijing Institute of Technology,Beijing 100081,China;School of Economics and Management,Beijing Information Science and Technology University,Beijing 100192,China;School of Public Policy and Management,University of Chinese Academy of Sciences,Beijing 100049,China)
出处
《统计与决策》
CSSCI
北大核心
2021年第20期166-170,共5页
Statistics & Decision
基金
北京市社会科学基金资助项目(18JDBGLB031)
北京市科委计划项目(Z191100010718001)。
关键词
企业创新绩效
零值数据
DEA
enterprise innovation performance
zero-value data
DEA
