Comfort versus Legitimacy? Student Perspectives on In-Course, Online, and AI-Mediated Assessment

This study examines how final-year undergraduate economics students evaluate contemporary assessment practices in a period of rapid digital change. Using survey data from a cohort of final-year undergraduate economics students at a UK university, it analyses perceptions of in-course assessment, online assessment, and generative AI in relation to fairness, inclusivity, learning support, stress, and procedural legitimacy. Rather than treating assessment attitudes as one-dimensional, the paper combines item-level evidence with a small number of transparent summary measures to distinguish between different sources of student support and concern. The findings reveal four main patterns. First, students evaluate in-course assessment positively overall, with especially strong support for its fairness and inclusivity. Second, this support is not unconditional: students value the reduction of final-exam stress through diversified assessment more strongly than they endorse the proportionality of in-course workload. Third, students view online assessment more positively in terms of comfort and convenience than in terms of fairness, integrity, and broader legitimacy. Fourth, generative AI is associated with substantial anxiety about penalties and uncertainty over acceptable use. Although exploratory, the results highlight three practical priorities for assessment design: preserving the fairness and inclusivity benefits of diversified assessment, improving workload calibration, and developing clearer governance around AI use. More broadly, the study shows the value of treating student views on assessment as multidimensional, especially in increasingly digital and AI-mediated educational environments.