基于数据包络分析的科研机构效率评价  

Performance Evaluation for Research Institutes Based on Data Envelopment Analysis

作　　者：李兆琼[1] 梁樑[1] 杨锋[1] 李勇军[1] 

机构地区：[1]中国科学技术大学管理学院,安徽合肥230026

出　　处：《系统工程》2010年第6期94-100,共7页Systems Engineering

基　　金：国家自然科学基金创新研究群体科学基金资助项目(70821001);国家自然科学基金资助项目(70801056)

摘　　要：为了更有效地配置有限的科研资源,有必要对科研机构的投入-产出效率进行研究。数据包络分析是评价相对效率的合适工具,但在直接应用于科研机构效率评价时面临两个难点:其一是科研机构的运作过程是一个复杂系统,其二是存在双属性变量。本文提出了一种针对科研机构的DEA效率评价方法,初步解决了这两个困难,能够比较合理的评价科研机构的效率。最后,对国内50所重点大学的效率评价的实证分析显示了本文方法的合理性。In order to allocate the limited research resources more effectively,it＇s necessary to study the input-output efficiency of research institutes.Data Envelopment Analysis（DEA） is a suitable tool to be used in efficiency or performance evaluation.However,the traditional DEA approaches can＇t be directly used in performance evaluation of research institutes because of two difficulties.One is that the operation process of research institutes is a complex system,the other is the existence of dual-role variables.The current paper proposes a performance evaluation method for research institutes,and can overcome the above two problems.As a result,the proposed method can generate reasonable and fair efficient scores for research institutes.Finally,an empirical research on the efficiency evaluation of 50 universities shows that the method is reasonable.

关 键 词：数据包络分析 双属性变量 科研机构 评价 

分 类 号：N945[自然科学总论—系统科学]