Make peer review a criteria-based learning task

Why it matters

Peer assessment has a modest average learning benefit, while newer evidence shows that reviewers often benefit from instructional support in producing better feedback. Give the reviewer a small set of explicit criteria, require evidence from the work, and ask for one concrete improvement. This makes the review process more disciplined without pretending that a peer is an expert. After reviewing, use the same criteria on your own attempt as a transfer step; treat that last step as a practical extension, not as a separately proven effect.

Steps

1. Choose two to four criteria that define a good attempt.
2. For each important judgment, point to the exact feature in the work that triggered it.
3. State the strength or gap in terms of a criterion, then suggest one specific revision.
4. Apply the same criteria to your own attempt before accepting peer comments as authority.
5. Verify disputed technical or high-stakes claims with an expert or authoritative source.

Check your result

At least one useful comment contains criterion + evidence + specific change, and you can apply the same criterion to your own work.

Keep this limit in mind

• The supporting evidence comes from educational peer assessment. Peer comments can be confidently wrong, and stronger feedback-process quality does not guarantee deeper subject knowledge. In the 2020 synthesis, peer assessment did not significantly outperform self-assessment, so use another person's attempt when the comparison itself adds useful information rather than assuming a peer is always necessary.