Hierarchical Regression for Modeling the Effect of Learning Behavior on Blended Academic Assessment for College English

In the environment of online and offline blended teaching for college English, English academic assessment in higher vocational colleges also presents a mixed approaches of teacher assessment, student assessment and AI assessment. This study conducted a questionnaire survey among 130 vocational college students, aiming to explore the influence of learning behavior on college English academic assessment. The research shows that higher vocational students have a higher degree of satisfaction with the blended academic assessment in general, and the grades from high to low are teacher assessment, AI assessment and student assessment. This study constructs the influencing factor model of teacher assessment, student assessment and AI assessment through hierarchical multiple regression, and uses Bootstrap sampling method to test the mediating effect of the model.The results show that, among the five factors of learning behavior, learning frequency has a positive impact on three academic assessment approaches, task completion rate and information exchange have a positive impact on teacher assessment, and information acquisition has a positive impact on student assessment. This study provides meaningful implications for the design of blended academic assessment in college English.