Educational Data Mining Of The National College Entrance Examination And The High School Entrance Examination Based On Hierarchical Linear Model

Research of education data mining has been focusing on exploring factors that influence student performance,hoping to find ways to improve student achievement,lower dropout rate,improve education quality,and improve education policy etc.Statistical analysis is applied to study student performance in traditional education analysis.While data mining has developed and performed well in fields like economic,education authorities and schools see the advantages of education data mining.Education data mining provides better solutions to problems such as the predictive analysis of student performance.At present most of the independent variables used in the analysis are students,personal information and test scores.Current resear-ch does not study the influence of different abilities of subj ects on student value-added,and does not study the different impacts of school characteristics on student abilities.Based on the questions setting and the official analysis report of the National College Entrance Exam(NCEE)and the High School Entrance Exam(HSEE),this thesis establishes indicators that measure student capabilities of junior high school.Based on the linear regression model and hierarchical linear model,this thesis studied following questions:(1)how students' capabilities and school features influence students value-added in total score and the score of five subjects from HSEE to NCEE?(2)How school characteristics cause difference in the influence of student abilities on student value-added in total score?The results show that different abilities of students have different effects on student value-added.In different schools,the influence of students' abilities on value-added are different.Students that have the same score but with different abilities will have different value-added in different schools.This thesis finds that:1)different abilities of the subject have different effects on students' value-added.Most of student abilities have a significant impact on student value-added.Among student abilities,the ability of English memory can best reflect the potential of student value-added.2)Normally speaking,the higher the ranking of schools,the higher the value-added of student total scores.However,the influence of school ranking in the value-added of subj ects does not all follow this rule.3)In different ranking level of schools,the influence of student abilities has different impact on student value-added in total score.