基于多任务委托代理模型的论文计分规则研究  

Study on Scoring Rule of Articles Based on Multi-Task Principal-Agent Model

作　　者：王春雷[1] WANG Chunlei(School of Business,Guangxi University,Nanning 530004,China)

机构地区：[1]广西大学商学院,广西南宁530004

出　　处：《科学与管理》2020年第4期64-68,共5页Science and Management

基　　金：国家自然科学基金地区科学基金项目(71363006)。

摘　　要：将高校教师撰写不同等级的学术论文看作是完成不同的任务,构建多任务委托代理模型解释论文计分规则偏向于低等级期刊的原因。模型考虑了撰写不同等级论文的努力投入差异以及发表不同等级论文的不确定性差异。研究结果表明,高等级和低等级期刊论文的投入差异越大,论文发表的不确定性差异越大,高校教师的风险厌恶程度越高,科研管理部门制定的最优论文计分规则就越偏向于低等级期刊论文。因此,制定论文计分规则时需要了解科研主体的特征以及科研工作的特点,而不能仅仅考虑不同等级论文成果对提升高校科研业绩的差异。This paper recognizes that writing different grades of academic articles by university teachers can be seen as completing different tasks,constructs a model based on the multi-task principal-agent model to explain article scoring rule which biased to low-level journal articles.The model considers the different effort input to write different grades of articles and the uncertainty of publishing them.The results show that the greater the difference in the input of high-level and low-level journal articles,the greater the difference in the uncertainty of publication,the higher the risk aversion of teachers,the best article scoring rule determined by scientific research management departments is more biased to low-level articles.Therefore,when making scoring rules,we need to understand the characteristics of scientific researchers and work,rather than just consider the differences between different grades of articles to improve the scientific research performance of universities.

关 键 词：委托代理模型 学术论文 计分规则 激励 

分 类 号：G311[文化科学]