科研管理

科研管理 ›› 2016, Vol. 37 ›› Issue (5): 122-131.

• 论文 • 上一篇    下一篇

变异系数加权的组合赋权模型及科技评价实证

石宝峰1，程砚秋2，王静1   

  1. 1 西北农林科技大学 经济管理学院，陕西 杨凌712100；
    2东北财经大学 会计学院，辽宁 大连116025
  • 收稿日期:2014-06-02 修回日期:2015-09-06 出版日期:2016-05-20 发布日期:2016-05-06
  • 通讯作者: 石宝峰
  • 基金资助:

    国家社会科学基金重大项目(06&ZD039，200701-201010)；国家自然科学基金面上项目(71171031，201201-201512；71373207，201401-201712)；国家自然科学基金青年项目(71201018,201301-201512；71503199，201601-201812)；中国博士后科学基金面上资助项目(2015M572608，201501-201612)。

A combination empowerment model based on variation coefficient weighted and its empirical study of S&T evaluation

Shi Baofeng1, Cheng Yanqiu2, Wang Jing1   

  1. 1. College of Economics & Management, Northwest A&F University, Yangling 712100, Shaanxi, China;
    2. School of Accounting, Dongbei University of Finance and Economics, Dalian 116025, Liaoning, China
  • Received:2014-06-02 Revised:2015-09-06 Online:2016-05-20 Published:2016-05-06

摘要： 通过R聚类剔除反映信息重复的指标，利用因子分析遴选对科技评价影响较大的指标，构建了反映科学发展观内涵的科技评价指标体系。在此基础上，利用变异系数对主观G1权重、客观余弦夹角权重的组合权重进行加权，建立了基于变异系数加权的组合赋权模型，并对中国14个典型的省份科技评价数据进行了实证。本文的创新与特色：一是通过变异系数对主观G1权重、客观余弦夹角权重的组合权重进行加权，构建了变异系数加权后组合权重最大的目标函数，进而确定了主、客观权重的组合系数θ1、θ2，反映了组合赋权中主、客观权重分布变异性越大、权重越大的赋权思路。弥补了现有组合赋权中主、客观权系数相等，无法有效反映专家知识经验和客观数据差异的不足。二是实证研究表明：科技投入不足是制约江西、广西两省科技发展较差的主要因素；科技对经济与社会的影响偏弱是制约四川省科技发展的关键因素；科技产出不足是制约黑龙江、山东、河南三省科技发展的瓶颈因素。

关键词: 组合赋权, 变异系数, 科技评价, G1赋权, 余弦夹角赋权

Abstract: The science and technology (S&T) evaluating index system of implicating the Scientific Development View Connotation is established, by using R cluster analysis to delete these indicators of reflecting repeated information, and using factor analysis to screen these indicators with the highest information content. On this basis, by combining the subjective G1 weight with the objective angle cosine weight utilizing the variation coefficient, the paper establishes a combination empowerment model based on variation coefficient weighted, and then carries out an empirical study of 14 typical provinces of China. The special and contributions of this paper lie on two aspects. Firstly, by combining the subjective G1 weight with the objective angle cosine weight utilizing the variation coefficient, the maximum objective function of combination weighting has been established. Then, the combination coefficient θ1 of the subjective G1 weight and the combination coefficient θ2 of the objective angle cosine weight can be calculated. At the same time, the empowering ideas that the bigger of the distribution variability between the subjective weighting and the objective weighting, the greater of the combination weighting, is reflected. Thus, the right time to make up the existing combination empowerment cannot effectively reflect the difference between the expert knowledge and the objective data, because the subjective weighting coefficient θ1 and the objective weighting coefficient θ2 is equal. Secondly, an empirical result show that the key factor of restricting Jiangxi and Guangxi S&T development is insufficient S&T input; the critical factor of constraining Sichuan S&T development is weak in science and technology's influence on the economic and social; the bottleneck element of restricting Heilongjiang, Shandong and Henan S&T development is insufficient S&T output.

Key words: combination empowerment, variation coefficient, S&T evaluation, G1 weighting, angle cosine weighting