Peer review is a long-standing classroom practice, and it can be very effective when students know what to look for. In practice, though, students often give vague praise or focus on spelling errors, especially when reviewing a long analytical essay on a dense novel like Doctor Zhivago. AI feedback, by contrast, is consistent and fast but lacks the human sense of an audience, so the two work best in combination.

A sensible sequence begins with student self-review, in which each writer checks their own draft against a short checklist. The next step is AI-generated feedback on structure, evidence, and clarity, which gives the student a baseline for revision. Peer review then focuses on what a human reader does best, such as noticing where an argument is confusing or unconvincing.

Giving peers a specific task improves the quality of their comments. Instead of asking for general feedback, assign each reviewer to answer three questions: what is the thesis, which piece of evidence is most convincing, and where did you get lost. These questions are easy to answer and produce useful information for the author.

Dividing the Work Sensibly

AI feedback is well suited to checking whether quotations are explained, whether paragraphs stay on topic, and whether the conclusion adds something new. Peers are better placed to comment on whether the essay is persuasive and engaging, since they experience it as real readers. Teachers then step in for the interpretive and conceptual points that neither tool nor peer can handle well.

• Self-check against a five-item rubric before submitting the draft
• AI feedback on structure, evidence use, and clarity
• Peer review focused on thesis, persuasiveness, and confusing passages
• Teacher comments on interpretation and depth of analysis
• A revision memo in which the student explains what changed and why

Training Students to Give Good Feedback

Students need practice giving peer feedback before they can do it well. A useful exercise is to have the whole class review an anonymous sample essay, compare comments, and discuss which were most helpful. This shows what specific, actionable feedback looks like and why comments such as "good job" do not help the writer.

Sentence starters can also help, such as "I was convinced when..." and "I wanted more evidence for..." These give reviewers a way to frame observations constructively. Over time, students begin to develop their own vocabulary for discussing writing.

Managing Trust in Different Sources of Feedback

Students sometimes trust automated feedback too much or too little. Explain that AI comments are suggestions to be weighed, not orders to be followed, and that they may occasionally miss the point of an unusual argument. Encouraging students to explain in their revision memo why they accepted or rejected a suggestion builds critical thinking.

The same applies to peer comments, which can be wrong or misleading. A writer who knows the novel well may recognize that a classmate has misread a scene. Teaching students to evaluate all feedback against the text is a skill that supports independent writing long after the class ends.

Grading the Revision Process

To make revision meaningful, include it in the grade. A modest portion of the essay score can reflect the quality of the revision and the reasoning in the memo. This rewards students who engage seriously with feedback, even if their first draft was rough.

Teachers should compare drafts to see whether the changes are substantial or superficial. Real revision involves rethinking arguments and reorganizing evidence, not just fixing typos. Highlighting examples of strong revision for the class reinforces the idea that writing improves through repeated, thoughtful effort.