Perceived Learning Support from Generative AI in Maritime Higher Education: A Socio-Technical Scaffolding Perspective

Generative artificial intelligence (GenAI) is increasingly adopted in higher education, yet evidence on its perceived learning support across cognitive, affective, and psychomotor domains remains limited, particularly in professional, socio-technical contexts. This study examined maritime students' perceptions of GenAI-supported learning using a socio-technical scaffolding framework grounded in Bloom's revised taxonomy. A cross-sectional survey of 250 students at a maritime university in China was conducted using a 17-item Likert-scale instrument. Exploratory factor analysis, Pearson correlations, and linear regression were conducted in IBM SPSS 27. EFA revealed two latent factors: cognitive-conceptual support and experiential-self-regulatory support. All three learning domains were strongly intercorrelated, and frequency of GenAI use significantly predicted perceived support across all domains. Lower-order cognitive support strongly predicted higher-order support. Students perceive GenAI as an integrated, multidimensional learning scaffold rather than a narrowly cognitive tool. These findings extend socio-technical scaffolding theory to professional maritime education and have implications for curriculum design.