Download PDFOpen PDF in browserHuman-in-the-Loop, Not Hands-Off: Do AI-Assisted TA Comments Improve Construction Writing Quality and Fairness in an Introductory BIM Course?10 pages•Published: June 2, 2026AbstractProfessors, Teaching Assistants, and Graders in construction management courses regularly evaluate open-ended project reports, a task requiring substantial time investment and subjective judgment. This study evaluates whether AI-assisted feedback where TAs draft comments using AI, then review and validate them, improves quality compared to traditional TA-only comments. Using quantity takeoff reports (N = 21 traditional, N = 12 AI-assisted), from introductory BIM course feedback was compared on three dimensions: specificity, actionability, and rubric alignment. AI-assisted comments scored higher across all dimensions, with strongest gains in actionability (d = 1.98, p = .09). Notably, these gains concentrated among junior and non-CM students, groups that received sparser feedback in baseline data. A reversal of equity gaps suggests the human-in-the-loop model may democratize feedback quality. We discuss ethical guardrails: human accountability (TAs own all comments), no automated grading, and transparency with students. Findings indicate AI-assisted feedback can enhance construction education assessment while maintaining pedagogical integrity and advancing fairness.Keyphrases: ai assisted feedback, construction management education, equity in assessment, rubric based evaluation In: Wesley Collins, Anthony Perrenoud and John Posillico (editors). Proceedings of Associated Schools of Construction 62nd Annual International Conference, vol 7, pages 226-235.
|

