Effect of Mentorship Training on Generative AI Adoption among Academic Staff in a Nigerian Health Sciences University: A Quasi-Experimental Study

Abstract

The rapid emergence of Generative Artificial Intelligence (GenAI) has created a growing awareness-adoption gap in Nigerian higher education, where lecturers are broadly familiar with tools such as ChatGPT and Gemini but lack the structured training needed to integrate them competently and ethically into academic work. This study investigated the effect of a structured six-week mentorship training programme on GenAI competence, confidence, and adoption among academic staff of David Umahi Federal University of Health Sciences (DUFUHS), Uburu, Ebonyi State, Nigeria, and examined whether gender and academic rank moderate training outcomes. A quantitative quasi-experimental pre-test/post-test design was employed. Using multistage sampling, 250 academic staff were recruited from eight faculties, yielding 238 usable responses (response rate: 95.2%). Participants were assigned to an experimental group (n = 125) that received the mentorship intervention and a control group (n = 113) that did not. Data were collected using a validated, researcher-developed instrument, with a Cronbach's alpha reliability coefficient of 0.89. Descriptive statistics, paired-samples t-tests, independent-samples t-tests, and Analysis of Covariance (ANCOVA) were used for analysis. Pre-training, 82.4% of respondents were aware of ChatGPT, yet only 44.9% had actively used any GenAI tool, and a mere 11.8% had applied it for mentoring. Following the intervention, the experimental group recorded a composite mean competence gain of 2.04 points on a five-point scale (pre-test M = 2.11; post-test M = 4.15), with all paired-samples t-test comparisons significant at p < .001 (overall t = 38.47). Post-intervention, active GenAI usage rose from 44.9% to 91.2%. The largest competence gains were in Ethical AI Use (+2.26) and AI-Enabled Mentoring (+2.22). Independent-samples t-tests confirmed significantly higher post-test scores in the experimental group compared to controls across all domains (p < .001). ANCOVA revealed that gender did not significantly moderate outcomes (F = 2.29, p = .133), whereas academic rank was a significant moderator (F = 4.53, p = .005, η² = .10), with professors recording the highest adjusted post-test means. Structured mentorship training significantly and meaningfully enhanced GenAI competence, confidence, and adoption among academic staff in a Nigerian health sciences university, directly bridging the awareness-adoption gap. Training outcomes were equitable across gender but were moderated by academic rank. Universities and regulatory bodies such as the National Universities Commission (NUC) should institutionalize structured GenAI mentorship programmes as a standard component of faculty development and accreditation requirements. Future research should assess long-term retention of training outcomes, scale interventions across multiple Nigerian universities, and develop context-specific AI governance frameworks responsive to the infrastructural and sociocultural realities of Nigerian higher education.