Statistical Analysis For Academic Achievement Based On Models Of Panel Data

The analysis of students' academic achievement in college has an important reflection and promotion effect on the development of college education.Therefore,it is necessary to analyze the grade data of college students.Based on the analysis of university performance by multivariate statistical methods such as correlation coefficient analysis,canonical correlation analysis,principal component analysis and principal component regression analysis,the paper further explores the data structure and trend of performance data,and based on the theory of common change of panel data mean,the CUSUM ratio test statistic is used and the structural analysis of the students' mathematics of Hefei University of Technology is carried out by Monte Carlo simulation.This article is divided into five chapters,and the structure of the article is as follows:In the first chapter,the research significance of this paper is firstly introduced.The application and development of multivariate statistical analysis methods in college performance analysis are presented.At the same time,we introduce the research status of panel data mean common breaks at home and abroad,as well as its research status in the analysis of grade data of college students.The second chapter introduces the related theories of canonical correlation analysis,principal component analysis,principal component regression analysis and other statistical methods used in college performance analysis,and it also introduces the common breaks analysis of panel data mean and Monte Carlo simulation.In the third chapter,the relevant analysis and canonical correlation analysis and PCA are used to analyze the data of the eight semesters of the 14 th grade students of the pre-school education majors and the corresponding scores of the corresponding test scores and draw corresponding conclusions,which are that the comprehensive ability of students in middle school will have a profound impact on college achievement,and to a certain extent,the result of the senior high school entrance examination determine the learning performance in college and universities.In the fourth chapter,based on the common change point theory of panel data mean,the CUSUM ratio test statistic is used and the structural analysis for the students' scores of mathematics of Hefei University of Technology is analyzed by Monte Carlo simulation,and achieve some results,which are that there are two change points in the fifth and seventh semesters in the grade data of information major,and there are no change points in the data of mathematic major among the eight semesters.Then thischapter concludes with a reasonable explanation of this result.The fifth chapter is the conclusion,summarizes the main research contents of this paper,and summarizes the shortcomings of this paper and puts forward the prospect.