Codes: AC9E7LA04, AC9E7LY02, AC9E7LY06. Goal: test and revise an explanation for a specific audience without overstating evidence. Prepare: Text D and the student's Day 18 draft; optional invented contexts: 50 model seats, 100 paper squares, 20 numbered cards.
- Set a test · 2 min. “A reader should be able to answer without the author standing beside the page.” Agree on two questions: “What is the whole?” and “What is being compared?”
- Demonstrate · 5 min. Read a purposely weak line: “It is 36, so 36% is 36.” Ask which context is missing. Repair: “If the whole has 100 seats, 36% names 36 seats.” Keep the condition attached to the result.
- Plan feedback · 5 min. Partners choose roles: reader and listener/author; a private teacher route is equally valid. The reader points to a confusing word or missing label, not a person's ability. Authors note the exact question.
- Try/revise · 7 min. Use a learner route to test yesterday's explanation with a peer/teacher or an imagined reader using the two fixed questions. Revise an opening, caption or connective. Explain why the revision helps this audience.
- Conference · 4 min. Sample two revisions. Check whether a new shortcut creates an overclaim:
3:4is not automatically3/4of the whole;36and36%need a whole and units. Praise evidence-bound wording, not display polish. - Exit · 2 min. “My reader asked ; I changed ___ because .” Record actual reader test versus simulated self-check; both aid planning but are different evidence.