Unraveling the Multidimensional Mechanism of College Students' Satisfaction with AI-Assisted Courses: A Structural Equation Modeling Approach

Abstract

Artificial Intelligence (AI) has deeply penetrated the entire process of higher education, and the satisfaction of college students with artificial intelligence in courses is crucial. Currently, the mechanisms influencing university students' satisfaction with the application of Homo sapiens artificial intelligence in courses remain unclear. This study focuses on the artificial intelligence scenario of higher education courses, and conducts an empirical analysis of student satisfaction data(N=462) through structural equation model (SEM). In this study, we constructs a three-dimensional influence mechanism model including learning effectiveness, technical functionality and innovation, content accuracy and credibility. The research results indicate that: three dimensions all significantly positively affect student satisfaction, but there are obvious differences in the intensity and mechanism of the curriculum. Content accuracy is the core evaluation criterion (standardized path coefficients is 0.54), and the problem of knowledge illusion needs to be solved through the construction of discipline corpus and adversarial training mechanism. Learning efficiency is the key driving force, and AI emotional support systems trigger interest conversion through emotional regulation, but we need to be wary of the deprivation of control caused by algorithm black boxes. Technological innovation needs to be guided by content form innovation, enterprises should optimize multi-modal generation capabilities, and teachers need to reconstruct personalized teaching design. This study provides theoretical support and practical path for improving the reliability and satisfaction optimization of AI education applications.