Generative Artificial Intelligence (GenAI) is rapidly transforming higher education, yet its impact on learning experiences remains contested. Existing research often isolates either cognitive outcomes (e.g., comprehension, creativity) or affective outcomes (e.g., motivation, engagement), leaving a gap in integrated analyses that also account for heterogeneity across student groups. This study investigates both dimensions simultaneously by examining university students’ perceptions of GenAI, focusing on learning, creativity, motivation, and engagement, alongside perceived risks such as overreliance, ethical concerns, and difficulties in verifying accuracy. Data were collected from 93 students and analyzed through Spearman’s correlations and unsupervised clustering (k-means) with PCA visualization. Findings indicate low to moderate positive correlations between GenAI usage and learning outcomes, particularly problem-solving and motivation. Cluster analysis reveals diverse usage–perception profiles, including paradoxical cases where frequent users report limited cognitive benefit. These results align with Technology Acceptance Model (TAM) and UTAUT assumptions of perceived usefulness and performance expectancy, while also showing that digital literacy moderates these relationships, especially in critical thinking and responsible use. The study contributes by integrating cognitive and affective outcomes, revealing latent profiles beyond averages, and bridging adoption models with responsible AI frameworks. Practical implications highlight the need for AI literacy training, ethical policies, and instructional design to foster effective and responsible GenAI integration in higher education.
Otman et al. (2025) studied this question.