The power of beliefs: How Generative Artificial Intelligence (GenAI) influences academic performance

The application of Generative Artificial Intelligence (GenAI) in higher education is expanding rapidly. However, little research focusses on the influence of students' beliefs on the application effectiveness of GenAI and their academic performance. This study combines the Belief-Action-Outcome (BAO) model and the Multimotive Information System Continuance (MISC) model to develop a theoretical framework. A mixed-methods approach, specifically the explanatory sequential design, was employed. In the quantitative stage, 724 valid questionnaires from students with GenAI usage experience were collected and the structural equation modeling technology was used. In the qualitative stage, 15 students from diverse backgrounds are interviewed. Three valuable findings are identified. First, design expectation fit has a significant positive impact on beliefs. Second, beliefs have a significant positive influence on continuous use. Third, continuous use has a significant positive effect on task performance and technical performance. The theoretical contribution of this study is that the theoretical model constructed by the research can reflect the whole process of the influence of students' beliefs in new information technology on their academic performance, which can provide practical implications for optimizing GenAI applications in education.