一种新的科教组合评价方法--共性数据排序选择模型

Combined Evaluation of Science and Technology:An Application of Common Data Ordered Choice Model

为了解决科技教育评价中针对同一评价对象,选取相同的指标,采用同样的数据,但不同评价方法得出的评价结果不一致的问题,本文以《泰晤士报高等教育副刊》世界大学排名为例,提出了一种新的组合评价方法——共性数据排序选择模型,其原理是,首先用各种可行的评价方法对科技教育评价对象进行评价,然后将评价结果排序后进行分级,筛选出各种评价方法公认的评价对象,接着采用排序多元选择模型进行回归,得到各指标的回归系数,将其标准化后作为组合评价的权重,最后进行加权汇总得到评价结果。共性数据排序选择模型克服了其他组合方法少数服从多数,对评价方法有限制的缺点。

In order to solve the problem of different methods producing different results in S＆T evaluation. This paper uses a new method CDOC（Common Data Ordered Choice Model）based on the 2007 times higher-QS world university rankings. First different appropriate methods are used to assess the university rankings, then different results are classified and the common results are selected. At last the ordered choice model is used to analyze the common data; the coefficient is the indicators＇ weight after standardization. CDOC has no restrict with evaluation method.