Predicting Student Attrition in Higher Education through the Determinants of Learning Progress: A Structural Equation Modelling Approach

Abstract

Higher education policies are designed to facilitate students’ learning progression and academic success. Following Tinto’s integration theory and Bean’s attrition model, this study proposes a research model to investigate whether students prone to attrition can be pre-emptively identified through self-evaluating academic factors contributing to their learning progress. Theoretically, the learning progress is identified with student success, represented by factors amenable to intervention including the interaction with peers and instructors, teaching effectiveness, exam scores, absenteeism, students’ effort, and academic course-related variables. An exploratory and confirmatory factor analysis of 530 undergraduate students revealed that the indicators of learning progress in such students were channeled into two constructs. The results indicated that the teacher effectiveness and learning materials contributed most to the learning progress. Structural equation modelling revealed that the learning progress variables have a significant impact on students’ attrition status. A multi-group analysis confirmed the academic semesters to be a moderator in the mediating effects of the students’ grade point average (GPA). This model functions as a framework to design a student-oriented learning system promoting students’ learning experience and academic success.