| 摘要: |
| 文章考察了辛普森悖论对大学排名与综合评价方法的影响及其作用机制。研究发现辛普森悖论是客观存在的，处理不妥时会对排名科学性产生一些影响。指出可能会产生辛普森悖论的指标数据类型是商值型数据。在大学排名以及其他大多数综合评价实践中，会产生辛普森悖论的指标是少数的，对排名结果的影响一般也是有限的。对于有可能产生辛普森悖论的指标，一般是可以预先识别和预先做出有效调整的。同时，进一步研究了综合评价方法中可能存在的缺陷及其改进策略。 |
| 关键词: 辛普森悖论 商值型指标 大学排名 综合评价 |
| DOI： |
|
| 基金项目: |
|
| Can Simpson’s Paradox Fundamentally Deny Scientificity of University Rankings? — On Potential Faults in Comprehensive Assessment Methods and Improvement Strategies |
| TANG Jianmin |
| (Chinese Academy of Science and Education Evaluation，Hangzhou Dianzi University，Hangzhou, Zhejiang 310018) |
| Abstract: |
| This article studies influences of Simpson’s paradox on university rankings and comprehensive assessment methods as well as its mechanism. The research has found that, firstly, Simpson’s paradox exists objectively and will have impacts on scientificity of rankings when indices are not handled improperly. Secondly, Simpson’s paradox only occurs when there are ratio indices. Thirdly, problematical indices are rare and only have small impact on ranking results in most of university rankings and other comprehensive assessments. Fourthly, it is feasible to identify those problematical indices and then make adjustments. This study also finds potential faults in comprehensive assessment methods and proposes its improvement strategies. |
| Key words: Simpson’s paradox Ratio indices University rankings Comprehensive assessment |
