Artificial Intelligence Use as a Predictor of Student Creativity in Ugandan Higher Education

Steven Mbuga
Department of Humanities and Social Sciences
Faculty of Humanities and Social Sciences,
Victoria University, Kampala, Uganda
Email: smbuga@vu.ac.ug

Abstract: This study examined the relationship between artificial intelligence (AI) usage and student creativity among undergraduates at Victoria University, Kampala, Uganda, addressing a gap in literature on AI’s effect on creativity within private African higher education. An explanatory sequential mixed-methods design was used. From a target population of 300 undergraduates across three faculties, a stratified random sample of 169 was drawn, yielding 162 participants (N = 162). Data were collected via structured questionnaires and a divergent thinking task (Guilford’s Alternative Uses Test) under unassisted and AI-assisted conditions, plus 3 focus group discussions (n = 18) and key informant interviews (n = 5). Validity was confirmed by expert judgment (CVI = .88) and reliability by Cronbach’s alpha (α = .84) and inter-rater agreement (ICC = .89). Ethical approval came from the Victoria University Research Ethics Committee (VUREC), registered with UNCST. Data were analyzed using paired-samples t-tests, multiple linear regression (SPSS v28), and thematic analysis. AI assistance significantly increased fluency (d = 0.82 to 1.12, p < .001) and elaboration (d = 0.77 to 1.05, p < .001) but reduced originality in Humanities (t = -4.56, p < .001) and Business (t = -3.01, p = .004). Copy-paste habits negatively predicted originality (β = -.381, p < .001), while iterative prompting predicted it positively (β = .215, p = .008). Generative AI thus acts as a dual-edged catalyst, aiding ideation when used collaboratively but eroding originality when used passively. The study recommends discipline-specific AI literacy training and process-oriented assessment redesign at Victoria University.