摘要
为了提高高校科研管理绩效分析的科学性与准确性,采用随机前沿分析法(SFA)对高校科研绩效进行评估分析。该方法充分运用随机扰动项以及生产函数进行随机生产前沿的构造,再利用极大似然法进行逐个参数数值的估算,最后进行技术无效率项的条件期望当做技术效率值,能够将全部的决策单元作出准确的评价。实验证明,该方法采用极大似然估计法充分利用了科研绩效的投入和产出指标,精确度高,不易受到异常点的影响而使技术效率计算结果与实际情况偏差较大,具有一定的应用价值。
In order to improve the scientificalness and effectiveness of university research management performance analysis, stochastic frontier analysis was adopted to evaluate and analyze the university scientific research performance. Stochastic frontier analysis use production function and random disturbance term to construct the stochastic production frontier, it estimate the parameter values through maximum likelihood method. Then, take conditional expectation which is technical inefficiency as technical efficiency values to make it easy to evaluate all decision units. The experiment shows that this method take the full advantages of input and output indicators of research performance through maximum likelihood method. It has high accuracy and not easy to be affected by outliers to result in large deviation between calculation results and actual situation. It has great application value.
出处
《科技通报》
北大核心
2014年第1期220-223,共4页
Bulletin of Science and Technology
基金
军工保密纵向厅级项目(JUT213001)
南京邮电大学校级思想政治教育规划课题(XC213024)
江苏省高校哲学社会科学研究基金项目(2011SJD880135)
关键词
高校科研绩效
随机前沿分析法
因子分析法
university research management performance
stochastic frontier analysis
factor analysis
