Methods · Creation, transfer & long horizon · Short guide
F14
Peer-review rehearsal
Review someone else's work to sharpen your own standards: training the evaluative eye you'll need on your next draft.
When to use it
- Peer review is coming and you usually write vague niceties.
- You want sharper standards for your own work by practising on someone else's.
- You finish reviews quickly and suspect you are not using the rubric.
What to ask AI
Type something like: “Rubric: [ ]. Artifact to review (not mine): [ ]. I will write a peer review. Critique my review: specificity, rubric coverage, actionability. Do not review the artifact for me first.”
Ask for critique of your review, not a substitute review.
Keep the hard part on your side: you produce the evaluative language.
How it helps
Peer-review rehearsal trains the evaluative eye you need on your own work.
Giving feedback is practice in applying criteria, if you never practise, your standards stay vague.
The tell is giving great peer feedback while leaving the same flaws in yours.
How to run it
- Get a peer's draft (or a sample) and the real rubric.
- Write a full review yourself: specific, criterion-linked, actionable.
- Ask the AI to critique your review, not to replace it.
- Revise the review once for specificity.
- Turn the same standards on your own draft the same day.
Example exchange
I pasted the rubric, my peer's draft, and my review of it. Critique my review only: specificity, rubric coverage, and whether the comments are actionable.
You wrote clarify the argument with no pointer. Which paragraph, and which rubric line does it miss? Also, you never addressed the use of evidence criterion.
I will cite paragraph 4 and tie the comment to the Evidence band descriptors.
Good. Rewrite that comment yourself, then apply the same Evidence check to your own draft.
Copyable prompt
Rubric: [ ].
Artifact to review (not mine): [ ].
I will write a peer review. Critique my review: specificity, rubric coverage, actionability.
Do not review the artifact for me first.
The Tell
If you give great peer feedback and leave the same flaws in yours, you practised performance for others, not standards for yourself.
If the AI wrote the review and you only signed it, you skipped the training.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Peer review trains evaluative criteria when reviewers must apply shared rubrics. Benefits for the reviewer’s own work are a known peer-review teaching claim.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI critique of your review quality is a delivery aid. It does not replace practising evaluation yourself. Efficacy of AI-coached peer review is speculative.