Redesigning an Argument Unit After AI-Assisted Drafts
At Lakeview High, the 11th-grade argument-writing unit culminates in a 4–6 page essay. Students submit a draft, participate in one peer-review session, and revise. This year, 14 of 26 drafts included at least one paragraph that closely matched an AI chatbot's phrasing; 9 students admitted pasting. Peer-review comments were mostly "good job" or "I like it." Only 5 students made substantive revisions between draft and final. The team's rubric has six categories (claim, evidence, reasoning, organization, style, conventions) at four levels, but the descriptors use words like "insightful" and "sophisticated" without examples. The summative is in three weeks, and the department must decide whether to ban devices, redesign the process, or both. The school's acceptable-use policy is silent on generative AI. Two teachers want to return to handwritten in-class essays; another argues that would penalize students with fine-motor IEP accommodations. The team has one 90-minute PLC session to make decisions.
1. Analyze the data. What do the AI use, peer-review comments, and revision rates suggest about the unit's design rather than only about student integrity?
Read the worked answer
The three data points are symptoms of a process problem. Fourteen of twenty-six drafts containing AI-like paragraphs suggests the task is high-stakes, product-only, and easy to outsource; nine admissions suggest students do not see a clear line or a reason to keep the work their own. Peer comments ("good job," "I like it") suggest peer review was assigned without a protocol, modeling, or a rubric students could apply. Only five students made substantive revisions suggests revision was not taught, monitored, or rewarded; a single peer session is not a revision cycle. The vague rubric descriptors ("insightful," "sophisticated") make self-assessment and peer feedback difficult. A ban addresses access, not instruction; students would still not know how to revise. The unit needs observable checkpoints, a shared definition of revision, and an authentic audience or oral defense. Integrity rules matter, but the data point first to redesign.
2. Propose a redesigned writing process for the remaining three weeks that addresses authenticity and revision. Include specific structures.
Read the worked answer
Week 1: rebuild the task. Students choose a local arguable issue, submit a one-page proposal with three sources, and write one in-class paragraph on demand. Teach source evaluation and a claim-evidence-reasoning frame. Week 2: staged drafting. Require an annotated outline, a draft with version history, and a revision memo naming two global changes. Peer review uses a checklist: "Is the claim debatable?" "Is the evidence cited?" "Does the reasoning explain the evidence?" "Is a counterclaim addressed?" Train reviewers with a model paragraph. Week 3: final draft plus a short oral defense in which students explain one paragraph's choices and answer two questions. Allow authorized AI use only for brainstorming or grammar, with a disclosure note. Provide a handwritten or oral alternative only as an accommodation, not as punishment. This design creates multiple authentic checkpoints, makes revision visible, and gives students a reason to own the prose.
3. Rewrite two rubric descriptors to be observable and usable by students. Explain how the change supports revision.
Read the worked answer
Original: "The claim is insightful and sophisticated." Revised: "The claim takes a debatable position, is stated in one or two sentences, and is supported by at least three pieces of evidence, including one counterclaim that is acknowledged and answered." Original: "The writing is organized and polished." Revised: "Each body paragraph opens with a reason connected to the claim, presents evidence with a citation, and explains how the evidence supports the reason; transitions link paragraphs." The change supports revision because students can compare their drafts to observable criteria, peer reviewers can cite a specific descriptor, and teachers can calibrate scoring with anchor papers. "Insightful" invites impressionistic grading; a debatable claim and cited evidence can be seen. The rubric should be given at the start, used in peer review, and attached to the revision memo so students name which criterion they improved. This also reduces the temptation to rely on AI, because the criteria reward visible thinking and revision rather than polished surface.