An empirical study on the evaluation system of hospital scientific research based on principal component analysis / 中华医学科研管理杂志

ABSTRACT

Objective:To construct a scientific research evaluation model through principal component analysis, and to explore scientific research evaluation methods for hospitals.Methods:The professional title, educational background, positions and scientific research output information of the scientific research personnel in the First Hospital of Jilin University from 2019 to 2020 were collected. Delphi expert consultation was used to determine the assignment value of each variable, and use SPSS 21.0 software was used to build the principal component analysis model and conduct model verification.Results:The study collected a total of 1 882 researchers′ information. The KMO value of the validity test and the Bartlett sphere test meet the requirements of principal component analysis (KMO=0.731, P<0.05); the model obtained a total of 7 principal components. Among them, principal component 1 represents researchers who published SCI papers, applying for national, provincial and ministerial level scientific research projects, and their part-time positions in academic societies. The second principal component represents the status of applying for patents and publications, and the third principal component represents the status of the awards. The scores of scientific research output of researchers were summarized and sorted according to disciplines, according to which the neurology, endocrinology and metabolism, neurosurgery, general surgery and orthopedics ranked better. The model verification results found that researchers with senior professional titles and doctoral degrees had the highest median weighted comprehensive score( P<0.05), suggesting that scholars with higher professional title levels and higher education received higher comprehensive scientific research output scores. Conclusions:The scientific research evaluation model constructed by this study can provide scientific data reference for the hospital scientific research evaluation.