ScholarWorks@숙명여자대학교

초록

This study aimed to empirically identify significant school factors that predict highand low-performing middle schools. We imported the academic achievement level of 12,702 students enrolled in 236 middle schools from the 2019 National Assessment of Educational Achievement (NAEA) and the results of a survey of students, teachers, and principals. Considering the study’s complex situation and substantial data, which include more than 200 explanatory variables for each subject analysis, we employed the LASSO penalized regression techniques. Conclusively, we found that the characteristics of student sub-groups and regional variables had a significant influence on the difference between high- and low-performing middle schools. Significant predictors in all subjects were the number of students from multicultural families, number of students receiving public assistance, degree of parent participation, rate of students' comprehension of the curriculum, and student's willingness to improve academic achievement. These differences between highand low-performing middle schools imply that administrative and financial policies need to be more individually customized according to school conditions.

키워드