When Artificial Intelligence Meets Blended Learning: A Quasi-Experimental Study of Physics Achievement at the University of Khartoum, Sudan

Artificial intelligence (AI) has increasingly been integrated into blended learning environments to enhance instructional effectiveness in higher education. However, empirical evidence regarding its impact on student learning outcomes in science education remains limited in developing contexts. This study examines the effect of AI-based blended learning on the academic achievement of second-year physics students at the College of Education, University of Khartoum, Sudan. A quasi-experimental design was employed with 40 students assigned to an experimental group (n = 20) and a control group (n = 20). Data were collected via an achievement test and analyzed using SPSS. The findings reveal a statistically significant improvement in physics achievement for students exposed to the AI-based blended learning approach compared with those receiving conventional instruction (p < 0.05). These results indicate that integrating AI with blended learning can enhance learning effectiveness in science education. The study highlights the importance of strengthening institutional infrastructure and professional development to support the implementation of AI-supported blended learning in higher education.