Answers are grounded in this paper and cite their sources.
This paper reports preliminary findings from a redesigned capstone assessment in a first-year Management unit at an Australian university involving over 300 students. Rather than treating generative AI as a content-generation tool, the assessment positions it as a simulated interview partner within a three-phase architecture: a pre-interview strategy memo, an AI-conducted adaptive interview, and a post-interview reflective analysis. Grounded in Schön’s reflective practice framework and Biggs and Tang’s constructive alignment, the design surfaces the gap between how students plan to communicate and how they do. Preliminary thematic analysis of 21 student reflective submissions reveals four patterns: (a) the strategy memo functions as a metacognitive anchor, making the gap between intention and execution visible and diagnosable; (b) students consistently demonstrate framework knowledge but revert to habitual communication patterns under adaptive pressure; (c) the structured reflection phase produces sophisticated self-diagnostic reasoning that connects transcript evidence to communication theory; and (d) students who completed multiple simulation attempts used iterative practice as a developmental mechanism rather than a remedial exercise. These findings suggest that AI-simulated interviews, when embedded within a reflective assessment architecture, generate qualitatively richer professional learning than traditional scripted interview preparation.
No takes yet. Share an insight, caveat, or question.
Mitchell et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: