AI Literacy and Academic Performance among Non-STEM Students in KNUST: The Roles of LMS Quality and Blended-Learning Satisfaction

In the modern university, learning has shifted beyond the traditional classroom. Students now navigate a hybrid world where face-to-face instruction blends with digital platforms that host materials, facilitate discussions, and deliver assessments. This digital layer is often managed by a Learning Management System, a software hub that acts as the central nervous system for a course. For these systems to work well, they must be reliable, easy to use, and responsive. However, the technology itself is only half the equation. The other half is the student's ability to understand and interact with the tools at their disposal, a skill set increasingly defined as AI literacy. This concept goes beyond simply knowing how to type a prompt; it involves the technical, ethical, and practical confidence to use intelligent tools effectively. As universities in Ghana and around the world integrate these technologies, a critical question arises: does knowing how to use these digital tools actually help a student get better grades, or is the connection more complicated?

Researchers at the Kwame Nkrumah University of Science and Technology in Ghana set out to untangle this relationship, focusing specifically on students who are not studying science, technology, engineering, or mathematics. These non-STEM students often face different educational challenges and may not receive the same level of computational training as their peers in technical fields. The research team surveyed 432 students from the university's Teacher Education and Educational Innovation departments. They asked these students to rate their own understanding of artificial intelligence, how they perceived the quality of the university's digital learning platform, how satisfied they were with the blended learning experience, and how well they were performing academically. By gathering these perspectives, the researchers aimed to see how these different factors flowed into one another to shape a student's success.

The study revealed a clear and sequential path that leads to better academic results. It found that students with higher levels of AI literacy tended to perceive the university's digital learning system as being of higher quality. When students understand the technology, the system feels more usable and reliable to them. This perception of quality, in turn, led to greater satisfaction with the blended learning experience. When students felt the system was working well for them, they reported being more satisfied with their overall education. Finally, this satisfaction was a strong predictor of academic performance. In other words, the path to better grades for these students did not run directly from knowing about AI to getting an A. Instead, the knowledge of AI improved how they saw the system, which made them happier with their learning environment, and that happiness drove their academic achievement.

Crucially, the researchers found that the quality of the digital system alone did not directly improve grades. A well-built platform, without the student's engagement and satisfaction, was not enough to guarantee success. The system quality mattered only because it influenced how satisfied the students felt. Furthermore, the study showed that AI literacy did have a direct, though smaller, positive effect on academic performance, but the primary route to success was through the chain of perception and satisfaction. The data also ruled out the idea that the digital system's quality alone could bridge the gap between AI knowledge and grades; that specific shortcut did not exist in the results. The strongest link was the full journey: from AI knowledge to system perception, to satisfaction, and finally to performance.

This discovery suggests that for universities to help non-STEM students succeed in a digital age, they cannot simply invest in better software or expect students to figure out the technology on their own. The findings indicate that institutions must support students in developing their AI literacy so they can navigate these systems with confidence. When students feel competent with the tools, they view the systems more positively, which leads to a more satisfying learning experience and, ultimately, better academic outcomes. The research highlights that in a blended learning environment, the human element of satisfaction and the cognitive element of literacy are just as important as the technical infrastructure itself.