Peer review can improve essays dramatically, but only when students know what to look for. Left alone with a Père Goriot draft, many reviewers write "good job" or fix commas and call it done. Structured protocols change that, giving students specific questions and a role to play. Adding AI feedback as a second layer gives them a reference point to compare against their own observations.

Start with a checklist tied to your rubric. Ask reviewers to underline the thesis, circle the strongest piece of evidence, and write one sentence summarizing the writer's argument. If the summary does not match what the writer intended, that gap tells the writer something valuable. These tasks are concrete enough that even hesitant students can complete them.

Pair students deliberately. Matching a strong writer with a developing writer can help both, as long as the stronger student is guided to explain rather than to correct. Alternatively, group by topic so reviewers have enough knowledge to challenge the writer's interpretation of, say, Delphine's behavior. Either approach works better than random assignment.

Questions That Produce Useful Comments

Replace vague instructions with targeted questions. Ask where the reader got confused, which paragraph felt strongest and why, and what the writer could add to make the argument more convincing. When reviewers answer in complete sentences and cite specific lines, their comments become far more actionable than a generic note of praise.

• Where does the essay state its main claim, and is it debatable?
• Which quotation or scene best supports the claim, and does the writer explain it?
• Which paragraph feels most like summary instead of analysis?
• What counterargument could a skeptical reader raise?
• What one change would make this essay stronger?

Where AI Fits Into the Process

After peer review, students can run their revised draft through an AI feedback tool aligned with your rubric. The tool gives a quick read on thesis clarity, evidence, and analysis, which students compare with their peers' comments. Where the two agree, the writer knows what to fix first. Where they differ, the writer must think about which advice makes sense.

Be transparent about the purpose. The AI is a feedback source, not an author, and students should still write and revise in their own words. Setting that expectation early helps prevent overreliance and keeps the focus on learning. You can also ask students to submit a short reflection explaining which suggestions they took and why.

Grading the Review Process Itself

If you want students to take peer review seriously, assign a small grade to the quality of their feedback. A simple rubric can reward specific comments, references to the text, and constructive suggestions. This motivates careful reading and signals that reviewing is a skill worth developing alongside writing.

Collect a few examples of excellent peer comments to share with the class. Students learn quickly from seeing what good feedback looks like in practice. Over a semester, the quality of reviews improves noticeably, and the burden on you to catch every issue decreases because students are catching more of them first.