Twelve Minutes an Essay
Ms. Pham teaches two sections of eleventh-grade English (32 students each) and is three weeks from the quarter deadline, with two graded argumentative essays required. Her current cycle: collect drafts, mark every essay by hand (about 12 minutes each), return them, collect finals. A student survey last quarter found that 68% looked only at the score, and in one section only 9 of 32 students changed a claim between draft and final. Her comments are mostly about conventions - comma splices, capitalization - because those are fast to mark. She has one 45-minute planning period a day and no aide. Department policy allows one revision per essay. She is considering dropping the draft entirely to save time.
1. Use the numbers to diagnose the feedback system. What is the 12-minute marking cycle actually buying, and why are the comments mostly about conventions?
Read the worked answer
The arithmetic first: 64 essays × 12 minutes = 768 minutes = 12.8 hours per round. At 45 minutes of planning a day, that is about 17 planning periods, or three and a half weeks, for one round of one essay - with two graded essays due in three weeks. The current cycle is not merely tight; it is infeasible, and it will be paid for with turnaround time, sleep, or skipped instruction.
What 12 minutes buys: a score and a list of surface errors. What it does not buy is anything students use.
Why conventions dominate is structural, not accidental. A comma splice can be circled in two seconds and looks like thorough teaching. Judging whether a claim is arguable, whether the evidence actually supports it, and whether the reasoning links them takes several minutes of thought per essay - and cannot be done 64 times in the available hours. The teacher is marking what the clock permits, not what the standard requires.
The outcome data confirm the mismatch: 68% read only the score; 9 of 32 changed a claim. Feedback that is not read and not acted on has no instructional value regardless of how many hours it cost.
Dropping the draft is the wrong fix. It removes the only revision opportunity and will likely depress quality, while leaving the underlying time problem untouched.
2. Design a feedback cycle for one essay that fits her actual hours and targets argument substance rather than conventions. Be specific about who gives what feedback and when.
Read the worked answer
Principle: stop marking everything; make the feedback selective, in-class, and required to be used.
Before drafting (one class period):
- Review only a thesis-plus-two-evidence-bullets checkpoint, 30–45 seconds each. A whole class can be checked in one period while students work. This is where claim problems are cheapest to catch.
During drafting (one to two class periods):
- Whole-class minilesson on 4–5 anonymized samples drawn from the class itself, naming the pattern (e.g., "evidence dropped in without a link"). This replaces dozens of identical individual comments.
- Structured peer review with a single-trait checklist on claim and evidence only: "Underline the claim. Is it arguable? Does each body paragraph name the evidence and explain the link?" Trained with a modeled example first.
- Teacher conferences, 8–10 per day, 3–4 minutes each, only with students whose checkpoint flagged a problem or who request one. Everyone else is revising.
On the draft:
- Mark one trait (claim/evidence/reasoning) with a short coded margin note, plus at most two conventions marks. No rewriting sentences.
- Use a single-trait rubric so the grade communicates one thing.
On the final:
- Require a revision memo (3 sentences: what I changed, why, what I would still fix). Conventions get graded here, not on the draft, so the draft stays focused on argument.
Resulting load: 64 checkpoints at under a minute plus about 20 conferences, instead of 12.8 hours of marking - and the feedback lands where the standard lives.
3. What evidence would show, within the three weeks available, whether the redesign is working - and what would show it is not?
Read the worked answer
Evidence of success (collect in three weeks, mostly from work already produced):
- Substantive revision rate. Count students whose claim or evidence changed between the checkpoint and the final. Baseline was 9 of 32; a target of two-thirds would be a real shift. The revision memo makes this directly countable.
- Single-trait rubric results. Score only claim-and-evidence on drafts and finals. Improvement from draft to final on the same trait is the cleanest signal that feedback was used.
- Sample re-scoring. Have a colleague independently score six finals against the same rubric. Agreement indicates the rubric is doing the work; disagreement means the criteria are still vague.
- Student report. Two questions on an exit ticket: "What did you change after feedback?" and "Which piece of feedback did you use?" Students who cannot answer either did not engage with feedback, whatever the score.
- Time log. Track actual minutes per stage. If the new cycle consumes the same 12.8 hours, it has failed even if scores rise.
Evidence the redesign is not working:
- Revision memos that describe only spelling and punctuation fixes.
- No movement on the draft-to-final trait score.
- Conference attendance collapsing because students do not see the value.
- Scores rising on claim/evidence but final essays still losing points for conventions - which would signal that the final stage needs the conventions instruction, not that the redesign failed.
Caveat to state plainly: three weeks is one cycle with one class set. Any conclusion is provisional, so the team should treat these numbers as a first read, not a verdict.