[Demo] Lecturer Perspectives on EMI Assessment Practices at Nguyen Tat Thanh University: Explaining GPA Improvement

Abstract

English-Medium Instruction (EMI) is expanding rapidly in Vietnamese higher education, yet the reasons behind recent grade improvements remain unclear. Drawing on literature relating assessment reliability and language scaffolding to student performance-but rarely centering teacher perspectives-we interviewed six purposively selected lecturers using semi-structured interviews (three STEM, three non-STEM). A five-prompt guide solicited narratives on rubric use, marking consistency, language-content balance, feedback practices, and perceived grade validity. A five-prompt guide elicited narratives on rubric use, marking consistency, language–content balance, feedback practices, and perceived grade validity. Transcripts were analyzed inductively using reflexive thematic analysis (RTA) with double-coding; intercoder agreement on an overlap sample reached κ = .85. Two intersecting themes are identified: (1) informal peer-moderation circles that foster shared interpretation of rubric criteria and rein in leniency drift and (2) wider use of bilingual feedback scaffolds (parallel Vietnamese-English annotations, discipline-specific glossaries) which can reduce language-related content loss and enable students to revise more effectively. The contribution of this article to the EMI quality assurance debates is based on process-level evidence of these lecturer-driven practices, combined with archived cohort GPA shifts and offering a scalable, low-cost model for strengthening assessment literacy and feedback uptake in comparable contexts.