Peer review has a mixed reputation among English teachers, largely because students often give vague comments like "good job" or focus on spelling errors instead of ideas. When the text is as complex as Ba Jin's Family, weak peer feedback is especially unhelpful, since the real challenges involve interpretation and evidence. Even so, peer review remains valuable, and pairing it with AI-generated feedback can make both stronger.

The idea is to treat each type of feedback as serving a different purpose. Peers are good at reporting what they understood and where they got lost, which gives writers a sense of how a real reader experiences the paper. AI feedback is good at checking alignment with the rubric and flagging missing elements, which a classmate might not recognize.

A teacher who combines them gets multiple perspectives without having to read every draft personally. Students receive input quickly, revise before the final submission, and arrive at the teacher's desk with stronger essays. That saves grading time and improves learning.

A Three-Step Revision Cycle

A simple cycle starts with a rubric-based AI response to the draft, which students use to revise obvious gaps such as a missing thesis or unexplained quotation. Next, students exchange the revised drafts for structured peer review using a short checklist. Finally, they submit the essay, and the teacher reads with a clearer sense of how much revision has already occurred.

• Draft the essay and run it through rubric-based AI feedback
• Revise the thesis, evidence, and explanation using the feedback
• Exchange drafts with a peer who answers three targeted questions
• Write a short reflection explaining which comments were used and why
• Submit the final essay along with the reflection for teacher review

Training Students to Give Useful Peer Feedback

Peer reviewers need specific questions to answer. Asking "Where does the writer explain why Jue Xin acts as he does, and is the explanation convincing?" produces a far more useful response than "What did you like?" Teachers can model a good review on a sample paragraph before sending students off to work in pairs.

It also helps to limit the number of comments. Asking each reviewer to identify one strength, one confusing moment, and one suggestion for evidence keeps the feedback focused. Students are more likely to act on three clear points than on a dozen scattered ones.

What AI Feedback Should and Should Not Do

AI feedback is most useful when it is tied directly to a rubric the teacher has written and the students have seen. It can note that an essay claims Ba Jin criticizes Confucian tradition but provides only one example, or that a paragraph lacks explanation of its evidence. It should not rewrite the essay or supply a thesis, since that removes the learning.

Teachers should set expectations about how students use the tool, for instance requiring them to submit their own draft and revise it themselves. Making the process visible, through a reflection or a revision log, discourages misuse and encourages thoughtful engagement. The teacher's role shifts from first-line editor to coach.

Measuring Whether the Process Works

A simple way to evaluate the cycle is to compare first drafts and final submissions on a single rubric row, such as use of evidence. If most students improve noticeably, the process is working. If improvement is uneven, the teacher can adjust the peer review questions or provide additional mini-lessons.

Over time, the process also changes how students think about writing. They begin to treat a draft as something to test and improve, not a finished product to hand in. In a unit built around a novel as layered as Family, that mindset leads to better essays and better discussions.