Three-stage hybrid workflow: API batch review, identification of doubtful cases, focused conversational teacher review¶

Strategy · All strategies
Evidence · no claims cited

Description¶

For large courses, the article proposes a combined workflow in three stages: (1) a first AI-assisted review via batch API processing generating structured feedback by rubric criterion; (2) identification of cases requiring attention, such as superficial or invalid assessments or evidence with access or interpretation difficulties; (3) focused teacher review through conversational interaction, verifying evidence in the original report and, when needed, directing the model's attention to a specific equation, table, graph, or fragment. This concentrates teacher time on aspects requiring disciplinary and pedagogical expertise.

Design Implications¶

Context¶

Requirements¶

• Requires instructions, rubric, and output format defined in advance for the API stage, and teacher involvement for the conversational stage.

Constraints¶

• API processing is less flexible for unanticipated situations; conversational interaction is less efficient for large numbers of reports and its responses depend on question formulation.

Target Learners¶

• higher education students in large experimental physics courses

Target Learning Goals¶

• formative feedback and assessment of laboratory reports

Affordances¶

• Evidence Present Versus Effectively Available Distinction

Related Strategies¶

• Use joint AI processing of report sets to identify recurring difficulties and guide collective feedback and teaching priorities
• Mitigate extraction errors by requiring high-resolution PDFs, equation-editor equations, and legible graph axes, units, and legends
• Use a hybrid AI-then-instructor grading workflow with immutable override history
• Implement AI-assisted report feedback through a four-stage phased pathway beginning with mature low-level tools
• Per-criterion guidance for instructors using AI feedback on lab reports — which rubric items to delegate, verify or review by hand

Examples¶

-

Key Sources¶

• Marcos Abreu, Cecilia Stari, Arturo C. Martí. (2026). AI-Assisted Assessment of Experimental Physics Laboratory Reports: Potential, Limitations, and Support for Teaching Practice. arXiv. https://arxiv.org/abs/2609.22417