This study aims to examine how artificial intelligence use (AIU) and personalized learning contribute to SDG-4 quality education (QE) and students’ mental health (MH) in higher education, and the study also investigates whether AIU statistically mediates the relationships between personalized learning and QE and MH.
A cross-sectional survey was conducted among 460 students at a Chinese university. The hypotheses were tested using partial least squares structural equation modeling.
Open learning was positively associated with both queality education and mental health, information management was positively associated with mental health but not with quality education, and knowledge creation was positively associated with quality education but not with mental health. None of the three personalized-learning dimensions were significantly associated with artificial intelligence use. In turn, artificial intelligence use was positively associated with both quality education and mental health. Artificial intelligence use showed significant indirect effects only between open learning and quality education and between open learning and mental health. No significant indirect effects were found for information management and, knowledge creation. Because the study used cross-sectional self-reported data, these findings indicate statistical associations rather than causal relationships.
The findings suggest that higher education institutions consider creating open and knowledge-rich learning environments, integrating AI tools that provide adaptive support and strengthening students’ information management (IM) skills. These factors were positively associated with quality education and mental health. However, the results should be interpreted cautiously because the study does not establish causal relationships or directly measure institutional progress toward SDG-4.
This study examines OL, IM and KC separately and tests their direct and indirect associations with QE and MH through AIU. The findings reveal dimension and outcome specific associations, with significant indirect effects only for the OL–QE and OL–MH relationships. Thus, AIU is not a common statistical pathway linking all personalized-learning dimensions with student outcomes.
