Skip to main content

AI-Enabled Feedback

Attribution

Original work: "Educators' guide to multimodal learning and Generative AI" — Tünde Varga-Atkins, Samuel Saunders, et al. (2024/25) — CC BY-NC 4.0
Adapted for UK Nursing Education by: Lincoln Gombedza, RN (LD)
Last Updated: December 2025

AI can be a powerful tool for formative feedback—giving students guidance on how to improve before they submit their final work.

PRIVACY WARNING (GDPR)

Educators: NEVER upload student work (assignments, reflections) into a public AI tool (ChatGPT, Claude, Gemini) unless you have a paid, "Zero-Data-Retention" enterprise agreement. Uploading student work without their explicit consent acts as a data breach.

1. The "Pre-Flight Check" (Student-Led)​

Encourage students to use AI to check their own work before submission. This frees up educators to focus on the content, not the grammar.

Prompt for Students:

"Act as an academic proofreader. Review my essay for:

  1. Clarity of argument.
  2. Passive voice usage.
  3. Logical flow. Do NOT rewrite the essay. Just give me a list of improvements."

2. Rubric Alignment​

Students can check if they have met the marking criteria.

Prompt for Students:

"Here is the marking rubric for a 'Distinction' level. Here is my draft introduction. Does my introduction meet the criteria? If not, what is missing?"

3. Feedback Generation (Enterprise Safe ONLY)​

If and only if your institution provides a secure, private AI instance (e.g., Microsoft Copilot with Commercial Data Protection): Educators can paste a specific section of a student's work to generate constructive feedback phrases.

Prompt:

"Suggest 3 constructive ways to phrase feedback for a student who has described a clinical procedure well but failed to reference the underlying evidence base."

Feedback vs. Marking​

  • AI is faster at checking syntax, structure, and breadth.
  • Humans are better at checking nuance, clinical safety, and "voice".
  • Rule: AI can Support the marker, but it should never Repace the marker.
Bias Check

AI models can be biased against non-standard English dialects. Ensure feedback does not unfairly penalize students relying on global English variations.