Abstract
This study examines the institutional capacity of Nigerian public universities to adapt artificial intelligence (AI) within the domain of science education research. Three hundred and sixty respond participated in the study. Drawing upon six hypotheses, the research analyzes the predictive and relational effects of digital infrastructure, academic staff digital competency, institutional policy frameworks, and research funding, alongside challenges relating to ethics and digital equity. A correlational survey design was employed using responses from 360 academic staff across science education, pure and applied science, educational technology/ICT in education, curriculum and instruction faculties. Descriptive statistics such as frequencies, and percentages were used to summarize the respondents’ demographic characteristics. Inferential statistics were employed to test the study hypotheses and determine relationships among variables: Pearson Product-Moment Correlation was used to examine the relationship between digital infrastructure, staff competency, governance, funding, and AI adaptation. Multiple Regression Analysis was used to predict the influence of institutional capacity factors on AI adaptation levels. Independent Samples t-Test was used to compare differences in AI capacity between federal and state universities. One-Way ANOVA was used to determine differences across institutional regions or types where applicable. Statistical significance was tested at p < 0.05. Findings revealed significant relationships between AI adaptation and digital infrastructure (r = 0.62), human capital (r = 0.58), institutional policy (r = 0.61), and research funding (r = 0.49), all at p < .001. An independent samples t-test showed a statistically significant difference in ethical and digital equity challenges between federal and state universities (t(358) = 3.414, p = .001). Multiple regression analysis further confirmed that institutional capacity variables jointly predicted 69.2% of the variance in AI adaptation levels (R² = 0.692, F(4, 225) = 125.43, p < .001), with infrastructure and human capital emerging as the strongest predictors. The study concludes that institutional disparities, particularly in infrastructure and ethical preparedness, hinder equitable AI integration across university types. It recommends targeted investments in digital infrastructure, strategic policy development, capacity-building for staff, and funding mechanisms to foster responsible and inclusive AI-driven science education research.