The distribution of scientific index weight based on information entropy and correlation coefficient

Research on Modern Higher Education

Rui Chi*¹ 

关键词:   scientific assessment;    weight coefficient;    amount of index information;    sample distribution;    index system;    assessment result;   

DOI  :  10.24104/rmhe/2017.02.02014

学科分类：社会科学、人文和艺术（综合）

来源: Asian Academic Press Co Limited

【 摘 要 】

The weights of scientific research indexes affect the accuracy of assessment. Information entropy reflects the amount of index information; correlation coefficients reflect the correlation between the indexes and assessment results. In this paper, it takes the normalized products of information entropy and correlation coefficients as the weights of scientific assessment indexes. The result of empirical analysis shows the reason why “The Core Journals” (native) and foreign journals being assigned with the largest weight coefficient is because both of its entropy and correlation coefficients are big. But some important index weights of the national projects and “A Journal Papers” are smaller because only a few people have the index data, and this will result in small index information entropy. It’s last verified by the sample distribution that the product of the scientific index and the weight coefficient is conducive to grading the assessment of the teachers.

【 授权许可】

CC BY-NC