Contents
- Refining Written Feedback for Clarity and Tone
- Generating Structured Feedback Aligned to a Rubric
- Creating Meaningful Whole-Cohort Feedback
- Supporting Audio or Video Feedback Production
- Using AI to Support Marker Standardisation (not marking)
- What Staff Explicitly Do Not Do (Boundary example)
- Making Practice Transparent to Students
1 – Refining Written Feedback for Clarity and Tone
Context
A marker has drafted personalised feedback for a 2,000‑word undergraduate essay. The feedback is accurate but dense and highly disciplinary.
How staff apply the guidance
The marker:
- Writes the feedback themselves, aligned to the marking criteria
- Inputs the draft feedback into an approved AI tool (e.g. Copilot)
AI is used to:
- Simplify sentence structure
- Improve clarity
- Adjust tone so it is constructive and supportive
Example prompt
“Rewrite the following feedback to be clearer and more supportive for a Level 4 student, without changing the academic judgement.”
Human judgement:
- The marker checks that the reworded feedback still reflects their judgement
- Removes or amends any phrasing that over‑generalises
- Ensures it remains specific to the student’s work
Alignment with Coventry University Group guidance:
- AI enhances communication, not judgement
- The academic remains fully responsible for feedback content
- No marks are generated or altered by AI
2 – Generating Structured Feedback aligned to a Rubric
Context
A module team uses a detailed rubric with multiple criteria. Markers have notes but want to ensure consistent alignment to each criterion in written feedback.
How staff apply the guidance
The marker:
- Assesses the work and decides the grade manually
- Writes brief notes against each rubric criterion
AI is used to:
- Turn bullet‑point notes into well‑structured feedback paragraphs under each criterion heading
Example prompt
“Convert these marker notes into coherent feedback paragraphs aligned to each assessment criterion. Do not add new judgement.”
Human judgement:
- The marker reviews each section for accuracy
- Adjusts emphasis where professional judgement is required
- Confirms that the overall feedback matches the awarded mark
Alignment with Coventry University Group guidance:
- AI supports consistency and structure
- Criteria are still applied by the academic
- Moderation processes remain unchanged
3 – Creating Meaningful Whole-Cohort Feedback
Context
After marking, the maker wants to provide cohort‑level feedback using Inspera, highlighting recurring strengths and development points.
How staff apply the guidance
The marker:
- Identifies common themes from their own marking (e.g. referencing issues, strong evaluation)
- Provides anonymised summary bullet points
AI is used to:
- Synthesise these points into a clear, student‑friendly cohort feedback statement
Example prompt
“Create a cohort feedback summary from these anonymised themes, suitable for Level 5 students.”
Human judgement:
- The marker checks that no individual student could be identified
- Ensures feedback aligns with what was actually seen in submissions
- Links feedback explicitly to future assessments (“feed‑forward”)
Alignment with Coventry University Group guidance:
- No individual grades are influenced
- Feedback supports learning enhancement
- Transparency is maintained
4 – Supporting Audio or Video Feedback Production
Context
A marker gives short, recorded feedback but wants a written transcript for accessibility and student preference.
How staff apply the guidance
The marker:
- Records feedback based on their own marking decisions
AI is used to:
- Transcribe the audio
- Lightly edit for clarity and readability
Human judgement:
- The marker reviews the transcript
- Corrects technical language or discipline‑specific terms
- Confirms consistency with marking decisions
Alignment with Coventry University Group guidance:
- Improves inclusivity and accessibility
- AI supports format conversion, not assessment
- Human oversight remains essential
5 – Using AI to Support Marker Standardisation (not marking)
Context
A team of markers wants to ensure consistent interpretation of criteria before marking begins.
How staff apply the guidance
The module team:
- Shares draft rubric descriptors and sample feedback
AI is used to:
- Suggest alternative wording
- Identify possible ambiguities in criteria language
Example prompt
“Highlight where these criteria might be interpreted differently by markers and suggest clearer wording.”
Human judgement:
- The team discusses and agrees final wording
- Ensures alignment with learning outcomes and PSRB expectations
Alignment with Coventry University Group guidance:
- AI supports assessment literacy
- Final decisions are academic and collective
- Used at design stage, not to assess students
6 – What Staff Explicitly Do Not Do (Boundary example)
Context
A marker considers uploading student assignments to a public AI tool to “get suggested marks”. This is not permitted practice.
Why:
- Marks must not be generated by AI
- Student data must not be uploaded to unapproved tools
- Doing so undermines academic integrity and student trust
Instead:
- The tutor marks manually
- Uses AI only after marking, for feedback enhancement or synthesis
7 – Making Practice Transparent to Students
Example wording in a module guide or Aula site
“Your work is marked by academic staff. AI tools may be used to help us structure or clarify feedback, but all academic judgement, marking decisions, and final feedback are made and approved by your tutors.”
This reinforces trust and aligns staff and student expectations.
