# Teacher-targeted worked key and next moves

**Public file, no access control.** The new cases are fresh relative to the lessons, not secure exams. Collect first responses before discussion; adapt numbers/contexts if learners may have read this key. Record access supports separately from substantive hints. All design records and methods are fictional.

## Daily look-fors

| Day | Expected evidence | Next move |
| --- | --- | --- |
| 11 | One paper tile, pattern family/border status; paper criterion separate from visitor accessibility. | Put `paper mark` and `person's experience` on different cards. |
| 12 | One maker/setting documented, several conditions absent; written repeat plan marked not observed. | Ask another group to follow a vague step; make setting/criterion operational. |
| 13 | Hatch18, Ring18, Solid12; pass25, redraw23, all48 in both directions. | Recount 14+9+2 and 4+9+10 before percentages. |
| 14 | 23/36≈63.9% selected versus 25/48≈52.1% complete; promotional card same author; no public use conclusion. | Keep correct subgroup arithmetic while repairing its label and inference. |
| 15 | See new Check A below. | Reteach the smallest missing component with another invented table. |
| 16 | Filter changes selected numerator/denominator, not source cells; all, two-family and single-family choices labelled. | Draw a frame around selected rows, recount, then remove the frame. |
| 17 | Two visual hierarchies compared from actual wording/placement; learner's own thumbnail revised from specific feedback. | Move full denominator and no-viewer-test note into the first read line. |
| 18 | Visual change, measurable paper criterion, written control and no invented result. | Split plan/result boxes and insist on `not observed`. |
| 19 | Actual two-panel artefact (when Arts claimed), full 25/48 first, selected 23/36 labelled, method limit and next paper check. | User-test only the first line against the source; revise overstatement. |
| 20 | See new Check B below. | Target denominator, omission or method claim separately. |

## Check A · worked first response

1. Rows: Grid `11+4=15`, Arc `7+5=12`, Stripe `2+6=8`; columns unbroken `11+7+2=20`, redraw `4+5+6=15`; full `15+12+8=35=20+15`.
2. Selected Grid+Arc unbroken `18/27=2/3≈66.7%`; all covers `20/35=4/7≈57.1%`. The statement mislabels the 27-cover subgroup as 35 covers. Stripe's eight records are omitted, including two unbroken.
3. Accept recording exact printer setting/paper stock/line criterion or asking a second blinded observer to code the **same** proof sheets and log disagreement. “Print more” without an operational condition is incomplete. The source measured no reader or access experience; one maker's inner-line mark is not a viewer outcome.
4. Accept: “In our invented one-setting paper check, 20 of 35 cover mock-ups had an unbroken inner line (about 57.1%); the method was not independently repeated and no readers tried them.” A separate clearly labelled 18/27 subgroup sentence may follow. Avoid claiming real usability.

**Response branch:** If all equals 27, restore Stripe. If “two-thirds” is called numerically false, keep 18/27 but relabel it. If reader access is asserted, ask where a human outcome appears in the source (nowhere). Preserve first-response evidence.

## Check B · worked first response

1. Full passes `8+6+1=15`, revisions `4+6+7=17`, all `32`; `15/32=46.875%≈46.9%`. Tall+Square passes `14`, total `(8+4)+(6+6)=24`; `14/24=7/12≈58.3%`. “Most” can describe **Tall+Square under that exact label**, not all 32.
2. Folded row, `1` pass and `7` revisions, is omitted. “Ready to hand to visitors” has no visitor evidence, safety or usability test. Even the full paper pass count would not justify it.
3. Accept: “In the fictional studio's one paper crop check, 15 of 32 mock-ups passed (46.9%). Do not claim visitor readiness; first document stock/setting and have a second reviewer repeat the predeclared crop check on labelled sheets.” A pause or revised paper test decision is defensible; no actual approval.
4. Operational control: name a fixed paper stock, recorded printer setting, identical crop margin and orientation, etc. Independent step: second reviewer uses the same predeclared rule and logs agreement/disagreement. A written plan is **not** an observed repeat.

**Response branch:** If 14/24 is read as all, frame Folded and recount. If the report hides the limit in tiny text, make it part of the title/first sentence. If a proposed retest is reported as completed, write `not observed` next to the imagined result.

## Evidence note

Record source/denominator, method/uncertainty, reasoning and audience/design choices separately. A graphic, tactile, AAC or text route can show the same claim evidence. A paper plan is planning evidence; an actual student-made two-panel visual artefact is needed for Visual Arts production evidence. This desk key has not been validated in a classroom.