Research On The Construction Of Scientific Data Quality Evaluation Index System In Colleges And Universities From The Perspective Of Responsible Subjects

At present,scientific research has entered the "fourth paradigm",that is,data intensive,the volume of scientific data is getting bigger and bigger,and the structure of scientific data is getting more and more complex,which makes the management of scientific data more difficult,the quality of scientific data is uneven,and seriously hinders the integration and deep repeated application of scientific data.Therefore,it is particularly critical to construct a reasonable evaluation index system of scientific data quality to evaluate scientific data and find problems and gaps.As the main force in the production and utilization of scientific data,universities need to strengthen the quality control of scientific data in the data-intensive scientific research environment,and build and improve the scientific data management structure and system of domestic universities.In addition,considering that the quality of scientific data is mostly controlled by human,the behavior of relevant subjects is the root cause of affecting the quality of scientific data,and quality is a kind of responsibility.Therefore,the university scientific data quality evaluation system constructed from the perspective of responsibility subject is more systematic and the results are more accurate.This study firstly reviews the theoretical and practical research on the evaluation of scientific data quality at home and abroad,makes a comparative analysis of the current research perspectives on the construction of the evaluation system of scientific data quality,and clarifies the advantages of the construction of the evaluation system from the perspective of the responsible subject.This study based on the data life cycle theory and the theory of total quality management,using literature research,network investigation,case analysis and induction,to select the sample(scientific journal "data policy text" data,scientific data center "data management policies and procedures" and associated documents "data quality assessment,management,control and share")as the research object,according to the definition of "scientific data quality responsibility subject",combing with the status quo of scientific data quality and using Wheeler model for reference,four types of scientific data quality responsibility subjects are determined,which are the main key responsibility subject,the main non-key responsibility subject,the secondary key responsibility subject and the secondary non-key responsibility subject.And the main key responsibility subjects are subdivided.Then,according to the 3W principle of "Who(responsible subject),When(scientific data life cycle stage)and What(scientific data quality responsibility)",the relevant quality responsibilities of four types of scientific data quality responsibility subjects in different stages are analyzed,extracted and summarized from the research samples.Then,on the premise of determining the system construction principles and evaluation dimensions,by theoretical and practical analysis,the evaluation indicator is extracted and determined from the responsibility determined by the subject of responsibility and various statistical information of the data set,and the university scientific data quality evaluation index system is preliminarily formed from the perspective of the subject of responsibility.The system is tested and adjusted by means of factor analysis according to the data of importance score of three-level index through questionnaire survey among experts and users.Through principal component analysis,the weight evaluation system is obtained.Finally,by using analysis of variance and T test,based on the customers "satisfaction" to score data,analyzes the different types of user satisfaction of scientific data quality difference,and through the IPA analysis,the importance and satisfaction of 40 indexes for the final score values through the contrast analysis of the two understand users’ perception about the quality of the scientific data used,the advantages and disadvantages of current scientific data quality are clarified.Based on this,countermeasures and suggestions to improve the quality of scientific data are put forward for the corresponding responsible subjects of scientific data quality.The innovation of this research lies in :(1)taking the subject of responsibility as the breakthrough point,the evaluation index of the corresponding dimension is extracted from the quality responsibility of the subject of scientific data quality,so as to construct the evaluation index system of university scientific data quality from the perspective of the subject of responsibility.(2)Based on the data life cycle theory and the five stages of the general process of the scientific data life cycle constructed in this study,the evaluation dimensions are determined,so as to shield the differences in evaluation contents caused by the diverse and interdisciplinary types of scientific data.(3)Empirical research on user satisfaction of scientific data quality is conducted based on IPA analysis method,and the advantages and disadvantages of current scientific data quality are obtained.Based on this,measures and suggestions on how to efficiently manage scientific data quality are provided from the aspects of production,process and output.