Wanjiru Kamau

Rubric-based feedback on a draft, with the top three fixes first

Scores a draft on six criteria with quoted evidence and a fix for each, ranks the three changes that matter most, and names the lines worth keeping.

At a glance

Best for: Writers of op-eds, essays, applications and blog posts who want honest, specific feedback they can act on instead of general praise.
Tested on: Claude · Opus 5.5
You fill in: draftpurpose
You get: Point: 4/5 Evidence: "It's time for Westfield to ban phones during the school day." Clear, but inconsistent: the headline says "in class," the… (full result below)

writing

editing

education

Prompt

Give me feedback on the draft below using this rubric. Be direct; I'd rather hear it now than from readers.

Draft:
{{draft}}

Purpose and reader: {{purpose}}

Score each criterion 1-5, with one quoted line from the draft as evidence and one specific fix:

1. Point: can a reader say what this argues after the first paragraph?

1. Evidence: are claims backed by examples, data or experience?

1. Structure: does each paragraph earn its place and follow from the last?

1. Reader fit: the right detail and tone for this reader?

1. Voice: does it sound like a person, and the same person throughout?

1. Ending: does it land, or just stop?

Then:

• Top 3 fixes, in the order that would improve the draft most.

• Keep: the 2 strongest lines, so I don't edit them out.

Don't rewrite the draft. You may rewrite at most one sentence per criterion to show what you mean.

Fill in

{{draft}}: The full drafte.g. Phones Don't Belong in Class, by Karen Holt...
{{purpose}}: What it's for and who will read ite.g. op-ed for the local paper, 250-word limit, read by parents and school board members

Example input

draft:
Phones Don't Belong in Class
By Karen Holt, 10th-grade English teacher, Westfield High

Everyone knows that phones are a distraction. Studies show that students can't focus when their phones are nearby, and teachers across the country agree. It's time for Westfield to ban phones during the school day.

Last spring, I tried an experiment. I put a pocket organizer by the door and asked students to leave their phones in it for the period. The first week was rough. By the third week, something changed. Students started talking to each other before the bell. One boy, who hadn't turned in an essay all year, finished a draft in class. A girl told me it was the first time she'd read a whole chapter without stopping.

Some parents worry about emergencies. But schools have phones in every office, and in an emergency, the last thing we need is 1,200 students calling their parents at once.

Technology is an important part of modern life, and students need to learn to use it responsibly. However, phones in class are simply not helping. The school board should act now.

We owe it to our kids.

purpose: op-ed for the local newspaper (250-word limit), read by parents and school board members ahead of an October board vote on a phone policy

2650

Generated Outputs (1)

2 hours ago

Claude

Opus 5.5

2 hours ago

Generated Output

1. Point: 4/5
Evidence: "It's time for Westfield to ban phones during the school day."
Clear, but inconsistent: the headline says "in class," the thesis says "the school day," and your experiment was one class period. Board members will vote on a specific policy, so pick one and use it throughout.

2. Evidence: 2/5
Evidence: "Studies show that students can't focus when their phones are nearby, and teachers across the country agree."
No study is named, and "teachers across the country agree" can't be checked. Your classroom experiment is the strongest evidence you have, but it has no numbers. Fix: name one study or cut the line, and add a count from your class (essays submitted before and after, say).

3. Structure: 3/5
Evidence: "Technology is an important part of modern life, and students need to learn to use it responsibly."
This paragraph concedes ground just before your ending and weakens it. Fix: replace it with the specific policy you want the board to adopt.

4. Reader fit: 3/5
Evidence: "the last thing we need is 1,200 students calling their parents at once."
Parents who worry about emergencies may read this as dismissive. Fix: tell them how they would reach their child: "Parents can always reach students through the front office, which is how emergencies should be handled anyway."

5. Voice: 3/5
Evidence: "Everyone knows that phones are a distraction."
The experiment paragraph sounds like you, a teacher who was there. The opening and closing sound like any op-ed. Fix: start with the experiment: "Last spring, I asked my 10th graders to leave their phones in a pocket organizer by the door."

6. Ending: 2/5
Evidence: "We owe it to our kids."
A stock line that could close any op-ed. Fix: end on the vote or on your strongest image.

Top 3 fixes

1. Open with the experiment and move the "studies show" line out of the lead (or cut it).

1. Say exactly what policy the board should vote for, and use the same scope in the headline and thesis.

1. Replace the stock ending with something only you could write.

Keep

• "One boy, who hadn't turned in an essay all year, finished a draft in class."

• "A girl told me it was the first time she'd read a whole chapter without stopping."

The body is about 180 words, so you have roughly 70 words of room for the policy detail and a number.

Why it works

A fixed rubric makes feedback specific and comparable instead of a general impression. Requiring a quoted line as evidence for every score stops vague criticism and shows exactly where the problem sits. One specific fix per criterion makes it actionable, and ranking the top 3 tells you where to spend limited editing time. The Keep list protects your best lines from over-editing. Capping rewrites at one sentence per criterion keeps it your draft: feedback, not a ghostwrite.

When not to use it

A rubric suits argument-driven writing; for fiction or poetry, ask for a reader's reaction instead. The scores are one reader's judgment, not a grade. If you need feedback on facts (whether a study says what you claim), check the source yourself; the model can flag unsupported claims but can't confirm them.

Comments (0)

No comments yet. Be the first to share your thoughts!