管理评论

›› 2019, Vol. 31 ›› Issue (5): 191-202.

• 组织与战略管理 • 上一篇    下一篇

基于模糊-Topsis的企业社会责任评价模型——以交通运输行业为例

孟斌1,2, 沈思祎1,2, 匡海波1,2, 李菲1, 丰昊月1   

  1. 1. 大连海事大学综合交通运输协同创新中心, 大连 116026;
    2. 大连海事大学航运经济与管理学院, 大连 116026
  • 收稿日期:2017-06-12 出版日期:2019-05-28 发布日期:2019-05-31
  • 通讯作者: 匡海波(通讯作者),大连海事大学航运经济与管理学院教授,博士生导师,博士。
  • 作者简介:孟斌,大连海事大学航运经济与管理学院讲师,硕士生导师,博士;沈思祎,大连海事大学航运经济与管理学院硕士研究生;李菲,大连海事大学综合交通运输协同创新中心硕士研究生;丰昊月,大连海事大学综合交通运输协同创新中心硕士研究生。
  • 基金资助:

    国家自然科学基金项目（71831002；71731003；71672016）；长江学者和创新团队发展计划（IRT_17R13）；辽宁省经济社会发展项目（2019lslktqn-021）；大连市社科联项目（2016dlskzd042；2017dlskzd034；2018dlskyb020）；大连海事大学教改项目（2018Y41）；中央高校基本科研业务费专项资金（3132019501；3132019502）。

Evaluating Model of Corporate Social Responsibility Based on Fuzzy Topsis: Taking Transportation Industry as Example

Meng Bin1,2, Shen Siyi1,2, Kuang Haibo1,2, Li Fei1, Feng Haoyue1   

  1. 1. Dalian Maritime University, Collaborative Innovation Center for Transport Studies, Dalian 116026;
    2. Dalian Maritime University, Shipping Economics and Management College, Dalian 116026
  • Received:2017-06-12 Online:2019-05-28 Published:2019-05-31

PDF (PC)

822

摘要：

本文以2015年42家交通运输行业上市企业为研究对象，以国际标准化组织ISO26000、全球报告倡议组织的G4标准为基础，通过主基底分析遴选出对企业社会责任评价结果影响显著的指标，采用相关分析剔除信息反映重复的指标，设立了包含6个一级准则层、12个二级准则层、39个指标的交通运输行业企业社会责任评价指标体系。通过模糊Topsis对指标进行赋权，构建交通运输行业企业社会责任绩效评价模型。文章的创新与特色：一是通过专家经验确定指标重要程度的最保守值、最可能值和最乐观值，通过三角模糊熵对指标进行赋权，保证了得到的权重更能真实的反映专家的主观意见，通过Topsis引入贴近度构造单个企业对于正理想解和负理想解的距离函数，测算企业社会责任绩效得分。二是通过Gram-Schmidt正交法对指标z-score标准化数据向量进行正交变换，并根据方差越大、相应指标携带的信息含量越多的思路，逐步筛选出最大方差对应的指标向量作为基底，直至新入选基底的方差达到阈值。通过保留已筛选出基底对应的指标并剔除剩余指标，建立了交通运输行业企业社会责任评价指标体系，确保筛选后的指标对结果有显著影响。

关键词: 企业社会责任评价, 交通运输行业, 主基底分析, 模糊Topsis

Abstract:

Based on ISO26000 of International Organization for Standardization and the G4 standard of the Global Reporting Initiative, this paper chooses 42 transportation industry listed companies as the research object, and uses Principal Basis Analysis to select the indicators that have significant impact on the evaluation of corporate social responsibility. And then, after using Correlation Analysis to eliminate the indicators that reflect repeated information, this paper builds an index system of social responsibility evaluation of transportation industry which include 7 first-level criterion layers, 12 second-level criterion layers and 39 indexes. The fuzzy Topsis is used to weight the index to construct the social responsibility performance evaluation model of the transportation enterprise. The innovation and characteristics of the study:First, this paper uses the expert experience to determine the most conservative value, the most likely value and the most optimistic value of index important degree. The triangular fuzzy entropy is also used in this paper to empower the index to ensure that the weight has the more realistic reflection on the subjective views of experts. Besides, this paper uses the close degree introduced by Topsis to construct the distance function of an individual firm to the positive ideal solutions and the negative ideal solution to calculate the score of the corporate social responsibility performance. Second, the Gram-Schmidt orthogonal method is used to transform the index z-score normalized data vector orthogonally, and according to the principle that the larger the variance is, the more information is carried in the corresponding index, this paper gradually selects the index vector corresponding to the maximum variance as the base until the new variance of the selected substrate reaches the threshold. By preserving the indexes selected by the base and removing the remaining indicators, the evaluation index system of corporate social responsibility in the transportation industry is constructed to ensure that the index after screening has a significant effect on the evaluation results.

Key words: corporate social responsibility evaluation, transportation industry, principal basis analysis, fuzzy Topsis