Artificial Intelligence, Learning Behavior and Critical Thinking in Higher Education-A PLS-SEM Mediation Approach

Artificial intelligence is transforming higher education through personalised learning, intelligent tutoring, adaptive systems, and generative tools. However, concerns remain that excessive reliance on artificial intelligence may encourage cognitive offloading and weaken independent reasoning. This study examines the relationships among artificial intelligence use, learning behaviour, and critical thinking among students at Africa University in Zimbabwe, with learning behaviour tested as a mediator. A quantitative cross-sectional survey design was used, and data from 299 undergraduate and postgraduate students were analysed using Partial Least Squares Structural Equation Modelling. The measurement model demonstrated strong reliability and validity, while the structural model supported all four hypotheses. Artificial intelligence use positively predicted learning behaviour and critical thinking, and learning behaviour strongly predicted critical thinking. Learning behaviour also partially mediated the relationship between artificial intelligence use and critical thinking. The findings indicate that the educational value of artificial intelligence depends less on the mere use of technology than on whether it promotes active, reflective, and self-regulated learning. The study contributes empirical evidence from a Sub-Saharan African university and provides implications for the responsible integration of artificial intelligence in higher education. Keywords—Artificial intelligence: Critical thinking: Higher education: Learning behaviour: Partial least squares structural equation modelling: Self-regulated learning