University Students’ Perceptions of Artificial Intelligence Usage and Its Perceived Impact on Critical Thinking Skills: Evidence from Somali Higher Education

Abstract

Artificial intelligence (AI) tools have rapidly penetrated higher education globally, yet their impact on students’ critical thinking skills remains a contested empirical question. This study presents findings from a quantitative cross-sectional survey examining university students’ perceptions of AI tool usage and its perceived effects on critical thinking skills in Somali higher education. Grounded in Cognitive Load Theory (Sweller, 1988), the Technology Acceptance Model (Davis, 1989), and Bloom’s revised taxonomy (Anderson & Krathwohl, 2001), a 67item structured questionnaire was administered to 105 undergraduate students at Hormuud University, Mogadishu, during Academic Year 2025–2026. Descriptive statistics, Pearson correlations, independent-samples t-tests, one-way ANOVA, and multiple regression were employed. Results reveal that 87.4% of students use ChatGPT, with 61.0% reporting daily or near-daily AI engagement. Critical thinking composite scores were moderate (M = 3.42, SD = 0.36 on a 5-point scale). Self-regulation emerged as the strongest predictor of perceived critical thinking development (r = 0.609, p < .001), substantially outweighing AI usage frequency (r = 0.222, p = .025). The multiple regression model (R² = .458, Adj. R² = .441) indicates that self-regulation, AI literacy, and usage frequency together explain 45.8% of variance in critical thinking perceptions. In alignment with Liu and Wang (2024), Wu and Yu (2024), and Brazão and Tinoca (2025), findings suggest that the quality of AI engagement is more influential than usage frequency for perceived cognitive outcomes. No significant gender or disciplinary differences were observed. Contextual findings highlight a moderate Englishlanguage barrier (M = 3.19) and inadequate institutional AI guidance (M = 2.76). This study makes a unique empirical contribution as one of the first studies to investigate AI and critical thinking in Somali higher education — a post-conflict, resource-constrained, and culturally distinctive context entirely absent from existing literature.