Artificial Intelligence in Online Assessment: A Review of Pedagogy, Support and Millennial Retention in Higher Education

Abstract

Artificial Intelligence (AI) now emerges as an engine in education, with online assessments feeling its most immediate reverberations. This systematic literature review examines the connection between AI tools and the dynamics of the labour market, focusing on scholars in Higher Education Institutions (HEIs). Anchored in Herzberg’s Two‑Factor Theory, the review interrogates how AI‑powered assessment platforms shape institutional support, retention tactics, and pedagogical practice. As Gen X declines in the labour market, millennials digital natives seeking innovative and supportive environments have become the prevailing cohort on campuses, pushing universities to rethink assessment practices. AI promises scalable, personalized, and streamlined feedback solutions, yet its rollout raises concerns about staff workload, institutional culture, and digital readiness. To investigate these tensions, the research team applied the (Preferred Reporting Items for Systematic Reviews and Meta‑Analyses) PRISMA systematic‑review protocol. A total of 1267 records were retrieved, with 477 duplicates excluded. 790 titles and abstracts were screened, of which 412 were excluded for irrelevance. 378 full‑text articles were assessed, and 292 were excluded for insufficient methodological rigor or lack of AI assessment focus. Finally, 86 studies were retained for synthesis The review connects retention, institutional backing, and generational transitions to the growing use of AI in assessment. Findings show that AI can lift achievement and reduce administrative burdens, but its promise depends on alignment with millennial values and strong institutional endorsement. This study enhances understanding of how AI can sharpen assessment methods while supporting talent retention. In closing, the paper outlines take‑aways for HEIs, acknowledges limitations, and offers directions for future research into AI’s pedagogical and workforce repercussions.

Ähnliche Arbeiten